Authors: Ardi Tampuu, Ilmar Uduste, Kristjan Roosild
In the behavioral cloning approach to end-to-end driving, a dataset of expert driving is collected and the model learns to guess what the expert would do in different situations. Situations are summarized in observations and the outputs are low or mid-level commands (e.g. brake, throttle, and steering; or trajectories). The models learn to match observations at time T to actions recorded at T or as simultaneously as possible. However, when deploying the models to the real world (or to an asynchronous simulation), the action predicted based on observations at time T gets applied at T + $\Delta$ T. In a variety of cases, $\Delta$ T can be considerable and significantly influence performance.
We first demonstrate that driving at two different speeds is effectively two different tasks. Delays partially cause this difference and linearly amplify it. Even without computational delays, actuator delays and slipping due to inertia result in the need to perform actions preemptively when driving fast. The function mapping observations to commands becomes different compared to slow driving. We experimentally show that models trained to drive fast cannot perform the seemingly easier task of driving slow and vice-versa. Good driving models may be judged to be poor due to testing them at "a safe low speed", a task they cannot perform.
Secondly, we show how to counteract the effect of delays in end-to-end networks by changing the target labels. This is in contrast to the approaches attempting to minimize the delays, i.e. the cause, not the effect. To exemplify the problems and solutions in the real world, we use 1:10 scale minicars with limited computing power, using behavioral cloning for end-to-end driving. Some of the ideas discussed here may be transferable to the wider context of self-driving, to vehicles with more compute power and end-to-mid or modular approaches.
Authors: Ruichen Zhang
When using a diffusion model for image editing, there are times when the modified image can differ greatly from the source. To address this, we apply a dual-guidance approach to maintain high fidelity to the original in areas that are not altered. First, we employ text-guided optimization, using text embeddings to direct latent space and classifier-free guidance. Second, we use perceptual similarity guidance, optimizing latent vectors with posterior sampling via Tweedie formula during the reverse process. This method ensures the realistic rendering of both the edited elements and the preservation of the unedited parts of the original image.
Authors: Litian Liang, Liuyu Bian, Caiwei Xiao, Jialin Zhang, Linghao Chen, Isabella Liu, Fanbo Xiang, Zhiao Huang, Hao Su
Building robots that can automate labor-intensive tasks has long been the core motivation behind the advancements in computer vision and the robotics community. Recent interest in leveraging 3D algorithms, particularly neural fields, has led to advancements in robot perception and physical understanding in manipulation scenarios. However, the real world's complexity poses significant challenges. To tackle these challenges, we present Robo360, a dataset that features robotic manipulation with a dense view coverage, which enables high-quality 3D neural representation learning, and a diverse set of objects with various physical and optical properties and facilitates research in various object manipulation and physical world modeling tasks. We confirm the effectiveness of our dataset using existing dynamic NeRF and evaluate its potential in learning multi-view policies. We hope that Robo360 can open new research directions yet to be explored at the intersection of understanding the physical world in 3D and robot control.
Authors: Ricardo Gonzalez, Peyman Nejat, Ashirbani Saha, Clinton J.V. Campbell, Andrew P. Norgan, Cynthia Lokker
Numerous machine learning (ML) models have been developed for breast cancer using various types of data. Successful external validation (EV) of ML models is important evidence of their generalizability. The aim of this systematic review was to assess the performance of externally validated ML models based on histopathology images for diagnosis, classification, prognosis, or treatment outcome prediction in female breast cancer. A systematic search of MEDLINE, EMBASE, CINAHL, IEEE, MICCAI, and SPIE conferences was performed for studies published between January 2010 and February 2022. The Prediction Model Risk of Bias Assessment Tool (PROBAST) was employed, and the results were narratively described. Of the 2011 non-duplicated citations, 8 journal articles and 2 conference proceedings met inclusion criteria. Three studies externally validated ML models for diagnosis, 4 for classification, 2 for prognosis, and 1 for both classification and prognosis. Most studies used Convolutional Neural Networks and one used logistic regression algorithms. For diagnostic/classification models, the most common performance metrics reported in the EV were accuracy and area under the curve, which were greater than 87% and 90%, respectively, using pathologists' annotations as ground truth. The hazard ratios in the EV of prognostic ML models were between 1.7 (95% CI, 1.2-2.6) and 1.8 (95% CI, 1.3-2.7) to predict distant disease-free survival; 1.91 (95% CI, 1.11-3.29) for recurrence, and between 0.09 (95% CI, 0.01-0.70) and 0.65 (95% CI, 0.43-0.98) for overall survival, using clinical data as ground truth. Despite EV being an important step before the clinical application of a ML model, it hasn't been performed routinely. The large variability in the training/validation datasets, methods, performance metrics, and reported information limited the comparison of the models and the analysis of their results (...)
Authors: Zaber Ibn Abdul Hakim, Najibul Haque Sarker, Rahul Pratap Singh, Bishmoy Paul, Ali Dabouei, Min Xu
A thorough comprehension of textual data is a fundamental element in multi-modal video analysis tasks. However, recent works have shown that the current models do not achieve a comprehensive understanding of the textual data during the training for the target downstream tasks. Orthogonal to the previous approaches to this limitation, we postulate that understanding the significance of the sentence components according to the target task can potentially enhance the performance of the models. Hence, we utilize the knowledge of a pre-trained large language model (LLM) to generate text samples from the original ones, targeting specific sentence components. We propose a weakly supervised importance estimation module to compute the relative importance of the components and utilize them to improve different video-language tasks. Through rigorous quantitative analysis, our proposed method exhibits significant improvement across several video-language tasks. In particular, our approach notably enhances video-text retrieval by a relative improvement of 8.3\% in video-to-text and 1.4\% in text-to-video retrieval over the baselines, in terms of R@1. Additionally, in video moment retrieval, average mAP shows a relative improvement ranging from 2.0\% to 13.7 \% across different baselines.
Authors: Amirhosein Chahe, Chenan Wang, Abhishek Jeyapratap, Kaidi Xu, Lifeng Zhou
This paper introduces an attacking mechanism to challenge the resilience of autonomous driving systems. Specifically, we manipulate the decision-making processes of an autonomous vehicle by dynamically displaying adversarial patches on a screen mounted on another moving vehicle. These patches are optimized to deceive the object detection models into misclassifying targeted objects, e.g., traffic signs. Such manipulation has significant implications for critical multi-vehicle interactions such as intersection crossing and lane changing, which are vital for safe and efficient autonomous driving systems. Particularly, we make four major contributions. First, we introduce a novel adversarial attack approach where the patch is not co-located with its target, enabling more versatile and stealthy attacks. Moreover, our method utilizes dynamic patches displayed on a screen, allowing for adaptive changes and movement, enhancing the flexibility and performance of the attack. To do so, we design a Screen Image Transformation Network (SIT-Net), which simulates environmental effects on the displayed images, narrowing the gap between simulated and real-world scenarios. Further, we integrate a positional loss term into the adversarial training process to increase the success rate of the dynamic attack. Finally, we shift the focus from merely attacking perceptual systems to influencing the decision-making algorithms of self-driving systems. Our experiments demonstrate the first successful implementation of such dynamic adversarial attacks in real-world autonomous driving scenarios, paving the way for advancements in the field of robust and secure autonomous driving.
Authors: Shuai Li, Minghan Li, Pengfei Wang, Lei Zhang
Recently, a few open-vocabulary methods have been proposed by employing a unified architecture to tackle generic segmentation and detection tasks. However, their performance still lags behind the task-specific models due to the conflict between different tasks, and their open-vocabulary capability is limited due to the inadequate use of CLIP. To address these challenges, we present a universal transformer-based framework, abbreviated as OpenSD, which utilizes the same architecture and network parameters to handle open-vocabulary segmentation and detection tasks. First, we introduce a decoder decoupled learning strategy to alleviate the semantic conflict between thing and staff categories so that each individual task can be learned more effectively under the same framework. Second, to better leverage CLIP for end-to-end segmentation and detection, we propose dual classifiers to handle the in-vocabulary domain and out-of-vocabulary domain, respectively. The text encoder is further trained to be region-aware for both thing and stuff categories through decoupled prompt learning, enabling them to filter out duplicated and low-quality predictions, which is important to end-to-end segmentation and detection. Extensive experiments are conducted on multiple datasets under various circumstances. The results demonstrate that OpenSD outperforms state-of-the-art open-vocabulary segmentation and detection methods in both closed- and open-vocabulary settings. Code is available at https://github.com/strongwolf/OpenSD
Authors: Zechuan Zhang, Zongxin Yang, Yi Yang
Creating high-quality 3D models of clothed humans from single images for real-world applications is crucial. Despite recent advancements, accurately reconstructing humans in complex poses or with loose clothing from in-the-wild images, along with predicting textures for unseen areas, remains a significant challenge. A key limitation of previous methods is their insufficient prior guidance in transitioning from 2D to 3D and in texture prediction. In response, we introduce SIFU (Side-view Conditioned Implicit Function for Real-world Usable Clothed Human Reconstruction), a novel approach combining a Side-view Decoupling Transformer with a 3D Consistent Texture Refinement pipeline.SIFU employs a cross-attention mechanism within the transformer, using SMPL-X normals as queries to effectively decouple side-view features in the process of mapping 2D features to 3D. This method not only improves the precision of the 3D models but also their robustness, especially when SMPL-X estimates are not perfect. Our texture refinement process leverages text-to-image diffusion-based prior to generate realistic and consistent textures for invisible views. Through extensive experiments, SIFU surpasses SOTA methods in both geometry and texture reconstruction, showcasing enhanced robustness in complex scenarios and achieving an unprecedented Chamfer and P2S measurement. Our approach extends to practical applications such as 3D printing and scene building, demonstrating its broad utility in real-world scenarios. Project page https://river-zhang.github.io/SIFU-projectpage/ .
Authors: Hongbin Lin, Juangui Xu, Qingfeng Xu, Zhengyu Hu, Handing Xu, Yunzhi Chen, Yongjun Hu, Zhenguo Nie
In the realm of 3D reconstruction from 2D images, a persisting challenge is to achieve high-precision reconstructions devoid of 3D Ground Truth data reliance. We present UNeR3D, a pioneering unsupervised methodology that sets a new standard for generating detailed 3D reconstructions solely from 2D views. Our model significantly cuts down the training costs tied to supervised approaches and introduces RGB coloration to 3D point clouds, enriching the visual experience. Employing an inverse distance weighting technique for color rendering, UNeR3D ensures seamless color transitions, enhancing visual fidelity. Our model's flexible architecture supports training with any number of views, and uniquely, it is not constrained by the number of views used during training when performing reconstructions. It can infer with an arbitrary count of views during inference, offering unparalleled versatility. Additionally, the model's continuous spatial input domain allows the generation of point clouds at any desired resolution, empowering the creation of high-resolution 3D RGB point clouds. We solidify the reconstruction process with a novel multi-view geometric loss and color loss, demonstrating that our model excels with single-view inputs and beyond, thus reshaping the paradigm of unsupervised learning in 3D vision. Our contributions signal a substantial leap forward in 3D vision, offering new horizons for content creation across diverse applications. Code is available at https://github.com/HongbinLin3589/UNeR3D.
Authors: Sunjae Yoon, Gwanhyeong Koo, Ji Woo Hong, Chang D. Yoo
Text-conditioned image editing has succeeded in various types of editing based on a diffusion framework. Unfortunately, this success did not carry over to a video, which continues to be challenging. Existing video editing systems are still limited to rigid-type editing such as style transfer and object overlay. To this end, this paper proposes Neutral Editing (NeuEdit) framework to enable complex non-rigid editing by changing the motion of a person/object in a video, which has never been attempted before. NeuEdit introduces a concept of `neutralization' that enhances a tuning-editing process of diffusion-based editing systems in a model-agnostic manner by leveraging input video and text without any other auxiliary aids (e.g., visual masks, video captions). Extensive experiments on numerous videos demonstrate adaptability and effectiveness of the NeuEdit framework. The website of our work is available here: https://neuedit.github.io
Authors: Mike Ranzinger, Greg Heinrich, Jan Kautz, Pavlo Molchanov
A handful of visual foundation models (VFMs) have recently emerged as the backbones for numerous downstream tasks. VFMs like CLIP, DINOv2, SAM are trained with distinct objectives, exhibiting unique characteristics for various downstream tasks. We find that despite their conceptual differences, these models can be effectively merged into a unified model through multi-teacher distillation. We name this approach AM-RADIO (Agglomerative Model -- Reduce All Domains Into One). This integrative approach not only surpasses the performance of individual teacher models but also amalgamates their distinctive features, such as zero-shot vision-language comprehension, detailed pixel-level understanding, and open vocabulary segmentation capabilities. In pursuit of the most hardware-efficient backbone, we evaluated numerous architectures in our multi-teacher distillation pipeline using the same training recipe. This led to the development of a novel architecture (E-RADIO) that exceeds the performance of its predecessors and is at least 7x faster than the teacher models. Our comprehensive benchmarking process covers downstream tasks including ImageNet classification, ADE20k semantic segmentation, COCO object detection and LLaVa-1.5 framework.
Code: https://github.com/NVlabs/RADIO
Authors: Zhipeng Bao, Yijun Li, Krishna Kumar Singh, Yu-Xiong Wang, Martial Hebert
Despite recent significant strides achieved by diffusion-based Text-to-Image (T2I) models, current systems are still less capable of ensuring decent compositional generation aligned with text prompts, particularly for the multi-object generation. This work illuminates the fundamental reasons for such misalignment, pinpointing issues related to low attention activation scores and mask overlaps. While previous research efforts have individually tackled these issues, we assert that a holistic approach is paramount. Thus, we propose two novel objectives, the Separate loss and the Enhance loss, that reduce object mask overlaps and maximize attention scores, respectively. Our method diverges from conventional test-time-adaptation techniques, focusing on finetuning critical parameters, which enhances scalability and generalizability. Comprehensive evaluations demonstrate the superior performance of our model in terms of image realism, text-image alignment, and adaptability, notably outperforming prominent baselines. Ultimately, this research paves the way for T2I diffusion models with enhanced compositional capacities and broader applicability. The project webpage is available at https://zpbao.github.io/projects/SepEn/.
Authors: Minye Wu, Zehao Wang, Georgios Kouros, Tinne Tuytelaars
Neural Radiance Fields (NeRF) revolutionize the realm of visual media by providing photorealistic Free-Viewpoint Video (FVV) experiences, offering viewers unparalleled immersion and interactivity. However, the technology's significant storage requirements and the computational complexity involved in generation and rendering currently limit its broader application. To close this gap, this paper presents Temporal Tri-Plane Radiance Fields (TeTriRF), a novel technology that significantly reduces the storage size for Free-Viewpoint Video (FVV) while maintaining low-cost generation and rendering. TeTriRF introduces a hybrid representation with tri-planes and voxel grids to support scaling up to long-duration sequences and scenes with complex motions or rapid changes. We propose a group training scheme tailored to achieving high training efficiency and yielding temporally consistent, low-entropy scene representations. Leveraging these properties of the representations, we introduce a compression pipeline with off-the-shelf video codecs, achieving an order of magnitude less storage size compared to the state-of-the-art. Our experiments demonstrate that TeTriRF can achieve competitive quality with a higher compression rate.
Authors: Xiao Zhang, David Yunis, Michael Maire
We present an approach for analyzing grouping information contained within a neural network's activations, permitting extraction of spatial layout and semantic segmentation from the behavior of large pre-trained vision models. Unlike prior work, our method conducts a wholistic analysis of a network's activation state, leveraging features from all layers and obviating the need to guess which part of the model contains relevant information. Motivated by classic spectral clustering, we formulate this analysis in terms of an optimization objective involving a set of affinity matrices, each formed by comparing features within a different layer. Solving this optimization problem using gradient descent allows our technique to scale from single images to dataset-level analysis, including, in the latter, both intra- and inter-image relationships. Analyzing a pre-trained generative transformer provides insight into the computational strategy learned by such models. Equating affinity with key-query similarity across attention layers yields eigenvectors encoding scene spatial layout, whereas defining affinity by value vector similarity yields eigenvectors encoding object identity. This result suggests that key and query vectors coordinate attentional information flow according to spatial proximity (a `where' pathway), while value vectors refine a semantic category representation (a `what' pathway).
Authors: Sahil Khose, Anisha Pal, Aayushi Agarwal, Deepanshi, Judy Hoffman, Prithvijit Chattopadhyay
Real-world aerial scene understanding is limited by a lack of datasets that contain densely annotated images curated under a diverse set of conditions. Due to inherent challenges in obtaining such images in controlled real-world settings, we present SkyScenes, a synthetic dataset of densely annotated aerial images captured from Unmanned Aerial Vehicle (UAV) perspectives. We carefully curate SkyScenes images from CARLA to comprehensively capture diversity across layout (urban and rural maps), weather conditions, times of day, pitch angles and altitudes with corresponding semantic, instance and depth annotations. Through our experiments using SkyScenes, we show that (1) Models trained on SkyScenes generalize well to different real-world scenarios, (2) augmenting training on real images with SkyScenes data can improve real-world performance, (3) controlled variations in SkyScenes can offer insights into how models respond to changes in viewpoint conditions, and (4) incorporating additional sensor modalities (depth) can improve aerial scene understanding.
Authors: Fangxun Shu, Lei Zhang, Hao Jiang, Cihang Xie
This paper presents Audio-Visual LLM, a Multimodal Large Language Model that takes both visual and auditory inputs for holistic video understanding. A key design is the modality-augmented training, which involves the integration of modality-specific tokens engineered to activate the appropriate visual and/or auditory encoder selectively. This mechanism is pivotal in enabling end-to-end joint training with video data at different modalities, including visual-only, audio-only, and audio-visual formats. Moreover, we introduce a high-quality video instruction dataset, derived from GPT-4. This dataset allows Audio-Visual LLM to adeptly process a variety of task-oriented video instructions, ranging from multi-turn conversations and audio-visual narratives to complex reasoning tasks. Extensive experiments demonstrate that Audio-Visual LLM impressively achieves strong zero-shot results across a range of video understanding tasks. For example, Audio-Visual LLM achieves an accuracy of 53.7% on MSRVTT-QA, outperforming non-LLM-based InterVideo by 6.6% and LLM-based Valley by 4.4%, respectively. Additionally, our Audio-Visual LLM also achieves competitive performance on audio tasks (e.g., AudioCaps).
Authors: Rahul Venkatesh, Honglin Chen, Kevin Feigelis, Khaled Jedoui, Klemen Kotar, Felix Binder, Wanhee Lee, Sherry Liu, Kevin A. Smith, Judith E. Fan, Daniel L. K. Yamins
The ability to understand physical dynamics is essential to learning agents acting in the world. This paper presents Counterfactual World Modeling (CWM), a candidate pure vision foundational model for physical dynamics understanding. CWM consists of three basic concepts. First, we propose a simple and powerful temporally-factored masking policy for masked prediction of video data, which encourages the model to learn disentangled representations of scene appearance and dynamics. Second, as a result of the factoring, CWM is capable of generating counterfactual next-frame predictions by manipulating a few patch embeddings to exert meaningful control over scene dynamics. Third, the counterfactual modeling capability enables the design of counterfactual queries to extract vision structures similar to keypoints, optical flows, and segmentations, which are useful for dynamics understanding. We show that zero-shot readouts of these structures extracted by the counterfactual queries attain competitive performance to prior methods on real-world datasets. Finally, we demonstrate that CWM achieves state-of-the-art performance on the challenging Physion benchmark for evaluating physical dynamics understanding.
Authors: Yi Chen, Yuying Ge, Yixiao Ge, Mingyu Ding, Bohao Li, Rui Wang, Ruifeng Xu, Ying Shan, Xihui Liu
Multimodal Large Language Models (MLLMs), building upon the powerful Large Language Models (LLMs) with exceptional reasoning and generalization capability, have opened up new avenues for embodied task planning. MLLMs excel in their ability to integrate diverse environmental inputs, such as real-time task progress, visual observations, and open-form language instructions, which are crucial for executable task planning. In this work, we introduce a benchmark with human annotations, EgoPlan-Bench, to quantitatively investigate the potential of MLLMs as embodied task planners in real-world scenarios. Our benchmark is distinguished by realistic tasks derived from real-world videos, a diverse set of actions involving interactions with hundreds of different objects, and complex visual observations from varied environments. We evaluate various open-source MLLMs, revealing that these models have not yet evolved into embodied planning generalists (even GPT-4V). We further construct an instruction-tuning dataset EgoPlan-IT from videos of human-object interactions, to facilitate the learning of high-level task planning in intricate real-world situations. The experiment results demonstrate that the model tuned on EgoPlan-IT not only significantly improves performance on our benchmark, but also effectively acts as embodied planner in simulations.
Authors: Qirui Yang, cheng qihua, Huanjing Yue, Le Zhang, Yihao Liu, Jingyu Yang
Raw low light image enhancement (LLIE) has achieved much better performance than the sRGB domain enhancement methods due to the merits of raw data. However, the ambiguity between noisy to clean and raw to sRGB mappings may mislead the single-stage enhancement networks. The two-stage networks avoid ambiguity by decoupling the two mappings but usually have large computing complexity. To solve this problem, we propose a single-stage network empowered by Feature Domain Adaptation (FDA) to decouple the denoising and color mapping tasks in raw LLIE. The denoising encoder is supervised by the clean raw image, and then the denoised features are adapted for the color mapping task by an FDA module. We propose a Lineformer to serve as the FDA, which can well explore the global and local correlations with fewer line buffers (friendly to the line-based imaging process). During inference, the raw supervision branch is removed. In this way, our network combines the advantage of a two-stage enhancement process with the efficiency of single-stage inference. Experiments on four benchmark datasets demonstrate that our method achieves state-of-the-art performance with fewer computing costs (60\% FLOPs of the two-stage method DNF). \textit{Our codes will be released after the acceptance of this work.}
Authors: Zehuan Huang, Hao Wen, Junting Dong, Yaohui Wang, Yangguang Li, Xinyuan Chen, Yan-Pei Cao, Ding Liang, Yu Qiao, Bo Dai, Lu Sheng
Generating multiview images from a single view facilitates the rapid generation of a 3D mesh conditioned on a single image. Recent methods that introduce 3D global representation into diffusion models have shown the potential to generate consistent multiviews, but they have reduced generation speed and face challenges in maintaining generalizability and quality. To address this issue, we propose EpiDiff, a localized interactive multiview diffusion model. At the core of the proposed approach is to insert a lightweight epipolar attention block into the frozen diffusion model, leveraging epipolar constraints to enable cross-view interaction among feature maps of neighboring views. The newly initialized 3D modeling module preserves the original feature distribution of the diffusion model, exhibiting compatibility with a variety of base diffusion models. Experiments show that EpiDiff generates 16 multiview images in just 12 seconds, and it surpasses previous methods in quality evaluation metrics, including PSNR, SSIM and LPIPS. Additionally, EpiDiff can generate a more diverse distribution of views, improving the reconstruction quality from generated multiviews. Please see our project page at https://huanngzh.github.io/EpiDiff/.
Authors: Lei Zhang, Fangxun Shu, Sucheng Ren, Bingchen Zhao, Hao Jiang, Cihang Xie
The massive growth of image-text data through web crawling inherently presents the challenge of variability in data quality. This paper introduces a novel algorithm, rooted in human knowledge, to compress this vast corpus of web-crawled image-text datasets to a compact and high-quality form. Our method unfolds in three major steps. First, we collect an image-text dataset, wherein each image is associated with multiple captions sourced from diverse origins. Then, to systemically capture human preferences regarding the best caption paired with each image, we establish a comprehensive set of both subjective and objective criteria for critically guiding the alignment assessment from labelers. Lastly, we train a reward model on the annotated dataset to internalize the nuanced human understanding of image-text alignment. The resulting reward model thus can act as a human-like referee to filter misaligned/low-quality image-text pairs. Extensive experiments demonstrate that we are able to secure (or even improve) model performance by compressing the image-text datasets up to ~90%. An impressive example is that, by aggressively reducing the total training sample from 130M to 15.5M (e.g., ~9x smaller), our BLIP-B/16 models still consistently show superior performance compared with the full-size-dataset counterpart on image-text retrieval (Flickr30K, COCO) by ~2.5% in Recall@1, and on image-captioning (Nocaps, COCO) by ~10.0% in CIDEr and ~2.7% in SPICE.
Authors: Nikhel Gupta, Zeeshan Hayder, Ray P. Norris, Minh Hyunh, Lars Petersson
We present a novel multimodal dataset developed by expert astronomers to automate the detection and localisation of multi-component extended radio galaxies and their corresponding infrared hosts. The dataset comprises 4,155 instances of galaxies in 2,800 images with both radio and infrared modalities. Each instance contains information on the extended radio galaxy class, its corresponding bounding box that encompasses all of its components, pixel-level segmentation mask, and the position of its corresponding infrared host galaxy. Our dataset is the first publicly accessible dataset that includes images from a highly sensitive radio telescope, infrared satellite, and instance-level annotations for their identification. We benchmark several object detection algorithms on the dataset and propose a novel multimodal approach to identify radio galaxies and the positions of infrared hosts simultaneously.
Authors: Tanveer Hannan, Md Mohaiminul Islam, Thomas Seidl, Gedas Bertasius
We present a novel end-to-end method for long-form video temporal grounding to locate specific moments described by natural language queries. Prior long-video methods for this task typically contain two stages: proposal selection and grounding regression. However, the proposal selection of these methods is disjoint from the grounding network and is not trained end-to-end, which limits the effectiveness of these methods. Moreover, these methods operate uniformly over the entire temporal window, which is suboptimal given redundant and irrelevant features in long videos. In contrast to these prior approaches, we introduce RGNet, a unified network designed for jointly selecting proposals from hour-long videos and locating moments specified by natural language queries within them. To achieve this, we redefine proposal selection as a video-text retrieval task, i.e., retrieving the correct candidate videos given a text query. The core component of RGNet is a unified cross-modal RG-Encoder that bridges the two stages with shared features and mutual optimization. The encoder strategically focuses on relevant time frames using a sparse sampling technique. RGNet outperforms previous methods, demonstrating state-of-the-art performance on long video temporal grounding datasets MAD and Ego4D. The code is released at https://github.com/Tanveer81/RGNet
Authors: Henry Hengyuan Zhao, Pan Zhou, Mike Zheng Shou
Large Language Models (LLMs) excel in understanding human instructions, driving the development of Multimodal LLMs (MLLMs) with instruction tuning. However, acquiring high-quality multimodal instruction tuning data poses a significant challenge. Previous approaches relying on GPT-4 for data generation proved expensive and exhibited unsatisfactory performance for certain tasks. To solve this, we present Genixer, an innovative data generation pipeline producing high-quality multimodal instruction tuning data for various tasks. Genixer collects datasets for ten prevalent multimodal tasks and designs instruction templates to transform these datasets into instruction-tuning data. It then trains pretrained MLLMs to generate task-specific instruction data and proposes an effective data filtering strategy to ensure high quality. To evaluate Genixer, a base MLLM model, Kakapo, is built and achieves SoTA performance in image captioning and visual question answering (VQA) tasks across multiple datasets. Experimental results show that filtered data from Genixer continually improves Kakapo for image captioning and VQA tasks. For the SoTA Shikra MLLM model on the image-region-related tasks, e.g., region caption and detection, Genixer also successfully generates corresponding data and improves its performance. Genixer opens avenues for generating high-quality multimodal instruction data for diverse tasks, enabling innovative applications across domains. The code and models will be released soon.
Authors: Bin Yang, Patrick Pfreundschuh, Roland Siegwart, Marco Hutter, Peyman Moghadam, Vaishakh Patil
LiDAR Upsampling is a challenging task for the perception systems of robots and autonomous vehicles, due to the sparse and irregular structure of large-scale scene contexts. Recent works propose to solve this problem by converting LiDAR data from 3D Euclidean space into an image super-resolution problem in 2D image space. Although their methods can generate high-resolution range images with fine-grained details, the resulting 3D point clouds often blur out details and predict invalid points. In this paper, we propose TULIP, a new method to reconstruct high-resolution LiDAR point clouds from low-resolution LiDAR input. We also follow a range image-based approach but specifically modify the patch and window geometries of a Swin-Transformer-based network to better fit the characteristics of range images. We conducted several experiments on three different public real-world and simulated datasets. TULIP outperforms state-of-the-art methods in all relevant metrics and generates robust and more realistic point clouds than prior works.
Authors: Demin Yu, Xutao Li, Yunming Ye, Baoquan Zhang, Chuyao Luo, Kuai Dai, Rui Wang, Xunlai Chen
Precipitation nowcasting is an important spatio-temporal prediction task to predict the radar echoes sequences based on current observations, which can serve both meteorological science and smart city applications. Due to the chaotic evolution nature of the precipitation systems, it is a very challenging problem. Previous studies address the problem either from the perspectives of deterministic modeling or probabilistic modeling. However, their predictions suffer from the blurry, high-value echoes fading away and position inaccurate issues. The root reason of these issues is that the chaotic evolutionary precipitation systems are not appropriately modeled. Inspired by the nature of the systems, we propose to decompose and model them from the perspective of global deterministic motion and local stochastic variations with residual mechanism. A unified and flexible framework that can equip any type of spatio-temporal models is proposed based on residual diffusion, which effectively tackles the shortcomings of previous methods. Extensive experimental results on four publicly available radar datasets demonstrate the effectiveness and superiority of the proposed framework, compared to state-of-the-art techniques. Our code will be made publicly available soon.
Authors: Balakrishnan Varadarajan, Bilge Soran, Forrest Iandola, Xiaoyu Xiang, Yunyang Xiong, Lemeng Wu, Chenchen Zhu, Raghuraman Krishnamoorthi, Vikas Chandra
Segment Anything Model (SAM) is a foundation model for interactive segmentation, and it has catalyzed major advances in generative AI, computational photography, and medical imaging. This model takes in an arbitrary user input and provides segmentation masks of the corresponding objects. It is our goal to develop a version of SAM that is appropriate for use in a photography app. The original SAM model has a few challenges in this setting. First, original SAM a 600 million parameter based on ViT-H, and its high computational cost and large model size that are not suitable for todays mobile hardware. We address this by proposing the SqueezeSAM model architecture, which is 50x faster and 100x smaller than SAM. Next, when a user takes a photo on their phone, it might not occur to them to click on the image and get a mask. Our solution is to use salient object detection to generate the first few clicks. This produces an initial segmentation mask that the user can interactively edit. Finally, when a user clicks on an object, they typically expect all related pieces of the object to be segmented. For instance, if a user clicks on a person t-shirt in a photo, they expect the whole person to be segmented, but SAM typically segments just the t-shirt. We address this with a new data augmentation scheme, and the end result is that if the user clicks on a person holding a basketball, the person and the basketball are all segmented together.
Authors: Shufan Li, Harkanwar Singh, Aditya Grover
The ability to provide fine-grained control for generating and editing visual imagery has profound implications for computer vision and its applications. Previous works have explored extending controllability in two directions: instruction tuning with text-based prompts and multi-modal conditioning. However, these works make one or more unnatural assumptions on the number and/or type of modality inputs used to express controllability. We propose InstructAny2Pix, a flexible multi-modal instruction-following system that enables users to edit an input image using instructions involving audio, images, and text. InstructAny2Pix consists of three building blocks that facilitate this capability: a multi-modal encoder that encodes different modalities such as images and audio into a unified latent space, a diffusion model that learns to decode representations in this latent space into images, and a multi-modal LLM that can understand instructions involving multiple images and audio pieces and generate a conditional embedding of the desired output, which can be used by the diffusion decoder. Additionally, to facilitate training efficiency and improve generation quality, we include an additional refinement prior module that enhances the visual quality of LLM outputs. These designs are critical to the performance of our system. We demonstrate that our system can perform a series of novel instruction-guided editing tasks. The code is available at https://github.com/jacklishufan/InstructAny2Pix.git
Authors: Yuzhou Huang, Liangbin Xie, Xintao Wang, Ziyang Yuan, Xiaodong Cun, Yixiao Ge, Jiantao Zhou, Chao Dong, Rui Huang, Ruimao Zhang, Ying Shan
Current instruction-based editing methods, such as InstructPix2Pix, often fail to produce satisfactory results in complex scenarios due to their dependence on the simple CLIP text encoder in diffusion models. To rectify this, this paper introduces SmartEdit, a novel approach to instruction-based image editing that leverages Multimodal Large Language Models (MLLMs) to enhance their understanding and reasoning capabilities. However, direct integration of these elements still faces challenges in situations requiring complex reasoning. To mitigate this, we propose a Bidirectional Interaction Module that enables comprehensive bidirectional information interactions between the input image and the MLLM output. During training, we initially incorporate perception data to boost the perception and understanding capabilities of diffusion models. Subsequently, we demonstrate that a small amount of complex instruction editing data can effectively stimulate SmartEdit's editing capabilities for more complex instructions. We further construct a new evaluation dataset, Reason-Edit, specifically tailored for complex instruction-based image editing. Both quantitative and qualitative results on this evaluation dataset indicate that our SmartEdit surpasses previous methods, paving the way for the practical application of complex instruction-based image editing.
Authors: Simon Giebenhain, Tobias Kirschstein, Markos Georgopoulos, Martin Rünz, Lourdes Agapito, Matthias Nießner
We present Monocular Neural Parametric Head Models (MonoNPHM) for dynamic 3D head reconstructions from monocular RGB videos. To this end, we propose a latent appearance space that parameterizes a texture field on top of a neural parametric model. We constrain predicted color values to be correlated with the underlying geometry such that gradients from RGB effectively influence latent geometry codes during inverse rendering. To increase the representational capacity of our expression space, we augment our backward deformation field with hyper-dimensions, thus improving color and geometry representation in topologically challenging expressions. Using MonoNPHM as a learned prior, we approach the task of 3D head reconstruction using signed distance field based volumetric rendering. By numerically inverting our backward deformation field, we incorporated a landmark loss using facial anchor points that are closely tied to our canonical geometry representation. To evaluate the task of dynamic face reconstruction from monocular RGB videos we record 20 challenging Kinect sequences under casual conditions. MonoNPHM outperforms all baselines with a significant margin, and makes an important step towards easily accessible neural parametric face models through RGB tracking.
Authors: Hidenobu Matsuki, Riku Murai, Paul H.J. Kelly, Andrew J. Davison
We present the first application of 3D Gaussian Splatting to incremental 3D reconstruction using a single moving monocular or RGB-D camera. Our Simultaneous Localisation and Mapping (SLAM) method, which runs live at 3fps, utilises Gaussians as the only 3D representation, unifying the required representation for accurate, efficient tracking, mapping, and high-quality rendering. Several innovations are required to continuously reconstruct 3D scenes with high fidelity from a live camera. First, to move beyond the original 3DGS algorithm, which requires accurate poses from an offline Structure from Motion (SfM) system, we formulate camera tracking for 3DGS using direct optimisation against the 3D Gaussians, and show that this enables fast and robust tracking with a wide basin of convergence. Second, by utilising the explicit nature of the Gaussians, we introduce geometric verification and regularisation to handle the ambiguities occurring in incremental 3D dense reconstruction. Finally, we introduce a full SLAM system which not only achieves state-of-the-art results in novel view synthesis and trajectory estimation, but also reconstruction of tiny and even transparent objects.
Authors: Junbum Cha, Wooyoung Kang, Jonghwan Mun, Byungseok Roh
In Multimodal Large Language Models (MLLMs), a visual projector plays a crucial role in bridging pre-trained vision encoders with LLMs, enabling profound visual understanding while harnessing the LLMs' robust capabilities. Despite the importance of the visual projector, it has been relatively less explored. In this study, we first identify two essential projector properties: (i) flexibility in managing the number of visual tokens, crucial for MLLMs' overall efficiency, and (ii) preservation of local context from visual features, vital for spatial understanding. Based on these findings, we propose a novel projector design that is both flexible and locality-enhanced, effectively satisfying the two desirable properties. Additionally, we present comprehensive strategies to effectively utilize multiple and multifaceted instruction datasets. Through extensive experiments, we examine the impact of individual design choices. Finally, our proposed MLLM, Honeybee, remarkably outperforms previous state-of-the-art methods across various benchmarks, including MME, MMBench, SEED-Bench, and LLaVA-Bench, achieving significantly higher efficiency. Code and models are available at https://github.com/kakaobrain/honeybee.
Authors: Trung-Hieu Hoang, Mona Zehni, Huy Phan, Minh N. Do
Despite the promising performance of current 3D human pose estimation techniques, understanding and enhancing their generalization on challenging in-the-wild videos remain an open problem. In this work, we focus on the robustness of 2D-to-3D pose lifters. To this end, we develop two benchmark datasets, namely Human3.6M-C and HumanEva-I-C, to examine the robustness of video-based 3D pose lifters to a wide range of common video corruptions including temporary occlusion, motion blur, and pixel-level noise. We observe the poor generalization of state-of-the-art 3D pose lifters in the presence of corruption and establish two techniques to tackle this issue. First, we introduce Temporal Additive Gaussian Noise (TAGN) as a simple yet effective 2D input pose data augmentation. Additionally, to incorporate the confidence scores output by the 2D pose detectors, we design a confidence-aware convolution (CA-Conv) block. Extensively tested on corrupted videos, the proposed strategies consistently boost the robustness of 3D pose lifters and serve as new baselines for future research.
Authors: Shaobo Xia, Jun Yue, Kacper Kania, Leyuan Fang, Andrea Tagliasacchi, Kwang Moo Yi, Weiwei Sun
We propose a weakly supervised semantic segmentation method for point clouds that predicts "per-point" labels from just "whole-scene" annotations while achieving the performance of recent fully supervised approaches. Our core idea is to propagate the scene-level labels to each point in the point cloud by creating pseudo labels in a conservative way. Specifically, we over-segment point cloud features via unsupervised clustering and associate scene-level labels with clusters through bipartite matching, thus propagating scene labels only to the most relevant clusters, leaving the rest to be guided solely via unsupervised clustering. We empirically demonstrate that over-segmentation and bipartite assignment plays a crucial role. We evaluate our method on ScanNet and S3DIS datasets, outperforming state of the art, and demonstrate that we can achieve results comparable to fully supervised methods.
Authors: Lyes Saad Saoud, Zhenwei Niu, Atif Sultan, Lakmal Seneviratne, Irfan Hussain
This research presents ADOD, a novel approach to address domain generalization in underwater object detection. Our method enhances the model's ability to generalize across diverse and unseen domains, ensuring robustness in various underwater environments. The first key contribution is Residual Attention YOLOv3, a novel variant of the YOLOv3 framework empowered by residual attention modules. These modules enable the model to focus on informative features while suppressing background noise, leading to improved detection accuracy and adaptability to different domains. The second contribution is the attention-based domain classification module, vital during training. This module helps the model identify domain-specific information, facilitating the learning of domain-invariant features. Consequently, ADOD can generalize effectively to underwater environments with distinct visual characteristics. Extensive experiments on diverse underwater datasets demonstrate ADOD's superior performance compared to state-of-the-art domain generalization methods, particularly in challenging scenarios. The proposed model achieves exceptional detection performance in both seen and unseen domains, showcasing its effectiveness in handling domain shifts in underwater object detection tasks. ADOD represents a significant advancement in adaptive object detection, providing a promising solution for real-world applications in underwater environments. With the prevalence of domain shifts in such settings, the model's strong generalization ability becomes a valuable asset for practical underwater surveillance and marine research endeavors.
Authors: Aditya Murali, Deepak Alapatt, Pietro Mascagni, Armine Vardazaryan, Alain Garcia, Nariaki Okamoto, Didier Mutter, Nicolas Padoy
Recently, spatiotemporal graphs have emerged as a concise and elegant manner of representing video clips in an object-centric fashion, and have shown to be useful for downstream tasks such as action recognition. In this work, we investigate the use of latent spatiotemporal graphs to represent a surgical video in terms of the constituent anatomical structures and tools and their evolving properties over time. To build the graphs, we first predict frame-wise graphs using a pre-trained model, then add temporal edges between nodes based on spatial coherence and visual and semantic similarity. Unlike previous approaches, we incorporate long-term temporal edges in our graphs to better model the evolution of the surgical scene and increase robustness to temporary occlusions. We also introduce a novel graph-editing module that incorporates prior knowledge and temporal coherence to correct errors in the graph, enabling improved downstream task performance. Using our graph representations, we evaluate two downstream tasks, critical view of safety prediction and surgical phase recognition, obtaining strong results that demonstrate the quality and flexibility of the learned representations. Code is available at github.com/CAMMA-public/SurgLatentGraph.
Authors: Joel Shor, Hiro-o Yamano, Daisuke Tsurumaru, Yotami Intrator, Hiroki Kayama, Joe Ledsam, Atsushi Hamabe, Koji Ando, Mitsuhiko Ota, Haruei Ogino, Hiroshi Nakase, Kaho Kobayashi, Eiji Oki, Roman Goldenberg, Ehud Rivlin, Ichiro Takemasa
$\textbf{Background and aims}$: Artificial Intelligence (AI) Computer-Aided Detection (CADe) is commonly used for polyp detection, but data seen in clinical settings can differ from model training. Few studies evaluate how well CADe detectors perform on colonoscopies from countries not seen during training, and none are able to evaluate performance without collecting expensive and time-intensive labels.
$\textbf{Methods}$: We trained a CADe polyp detector on Israeli colonoscopy videos (5004 videos, 1106 hours) and evaluated on Japanese videos (354 videos, 128 hours) by measuring the True Positive Rate (TPR) versus false alarms per minute (FAPM). We introduce a colonoscopy dissimilarity measure called "MAsked mediCal Embedding Distance" (MACE) to quantify differences between colonoscopies, without labels. We evaluated CADe on all Japan videos and on those with the highest MACE.
$\textbf{Results}$: MACE correctly quantifies that narrow-band imaging (NBI) and chromoendoscopy (CE) frames are less similar to Israel data than Japan whitelight (bootstrapped z-test, |z| > 690, p < $10^{-8}$ for both). Despite differences in the data, CADe performance on Japan colonoscopies was non-inferior to Israel ones without additional training (TPR at 0.5 FAPM: 0.957 and 0.972 for Israel and Japan; TPR at 1.0 FAPM: 0.972 and 0.989 for Israel and Japan; superiority test t > 45.2, p < $10^{-8}$). Despite not being trained on NBI or CE, TPR on those subsets were non-inferior to Japan overall (non-inferiority test t > 47.3, p < $10^{-8}$, $\delta$ = 1.5% for both).
$\textbf{Conclusion}$: Differences that prevent CADe detectors from performing well in non-medical settings do not degrade the performance of our AI CADe polyp detector when applied to data from a new country. MACE can help medical AI models internationalize by identifying the most "dissimilar" data on which to evaluate models.
Authors: Silvano A. Bernabel, Sos S. Agaian
This paper tackles the intricate challenge of improving the quality of nighttime images under hazy and low-light conditions. Overcoming issues including nonuniform illumination glows, texture blurring, glow effects, color distortion, noise disturbance, and overall, low light have proven daunting. Despite the inherent difficulties, this paper introduces a pioneering solution named Nighttime Dehazing, Low-Light Enhancement, and Light Suppression (NDELS). NDELS utilizes a unique network that combines three essential processes to enhance visibility, brighten low-light regions, and effectively suppress glare from bright light sources. In contrast to limited progress in nighttime dehazing, unlike its daytime counterpart, NDELS presents a comprehensive and innovative approach. The efficacy of NDELS is rigorously validated through extensive comparisons with eight state-of-the-art algorithms across four diverse datasets. Experimental results showcase the superior performance of our method, demonstrating its outperformance in terms of overall image quality, including color and edge enhancement. Quantitative (PSNR, SSIM) and qualitative metrics (CLIPIQA, MANIQA, TRES), measure these results.
Authors: Travis Driver, Andrew Vaughan, Yang Cheng, Adnan Ansar, John Christian, Panagiotis Tsiotras
This paper proposes the incorporation of techniques from stereophotoclinometry (SPC) into a keypoint-based structure-from-motion (SfM) system to estimate the surface normal and albedo at detected landmarks to improve autonomous surface and shape characterization of small celestial bodies from in-situ imagery. In contrast to the current state-of-the-practice method for small body shape reconstruction, i.e., SPC, which relies on human-in-the-loop verification and high-fidelity a priori information to achieve accurate results, we forego the expensive maplet estimation step and instead leverage dense keypoint measurements and correspondences from an autonomous keypoint detection and matching method based on deep learning to provide the necessary photogrammetric constraints. Moreover, we develop a factor graph-based approach allowing for simultaneous optimization of the spacecraft's pose, landmark positions, Sun-relative direction, and surface normals and albedos via fusion of Sun sensor measurements and image keypoint measurements. The proposed framework is validated on real imagery of the Cornelia crater on Asteroid 4 Vesta, along with pose estimation and mapping comparison against an SPC reconstruction, where we demonstrate precise alignment to the SPC solution without relying on any a priori camera pose and topography information or humans-in-the-loop
Authors: Hangfei Lin, Li Miao, Amir Ziai
Few-shot image classification is the task of classifying unseen images to one of N mutually exclusive classes, using only a small number of training examples for each class. The limited availability of these examples (denoted as K) presents a significant challenge to classification accuracy in some cases. To address this, we have developed a method for augmenting the set of K with an addition set of A retrieved images. We call this system Retrieval-Augmented Few-shot Image Classification (RAFIC). Through a series of experiments, we demonstrate that RAFIC markedly improves performance of few-shot image classification across two challenging datasets. RAFIC consists of two main components: (a) a retrieval component which uses CLIP, LAION-5B, and faiss, in order to efficiently retrieve images similar to the supplied images, and (b) retrieval meta-learning, which learns to judiciously utilize the retrieved images. Code and data is available at github.com/amirziai/rafic.
Authors: Mengwei Ren, Wei Xiong, Jae Shin Yoon, Zhixin Shu, Jianming Zhang, HyunJoon Jung, Guido Gerig, He Zhang
Portrait harmonization aims to composite a subject into a new background, adjusting its lighting and color to ensure harmony with the background scene. Existing harmonization techniques often only focus on adjusting the global color and brightness of the foreground and ignore crucial illumination cues from the background such as apparent lighting direction, leading to unrealistic compositions. We introduce Relightful Harmonization, a lighting-aware diffusion model designed to seamlessly harmonize sophisticated lighting effect for the foreground portrait using any background image. Our approach unfolds in three stages. First, we introduce a lighting representation module that allows our diffusion model to encode lighting information from target image background. Second, we introduce an alignment network that aligns lighting features learned from image background with lighting features learned from panorama environment maps, which is a complete representation for scene illumination. Last, to further boost the photorealism of the proposed method, we introduce a novel data simulation pipeline that generates synthetic training pairs from a diverse range of natural images, which are used to refine the model. Our method outperforms existing benchmarks in visual fidelity and lighting coherence, showing superior generalization in real-world testing scenarios, highlighting its versatility and practicality.
Authors: Philipp V. Rouast
This report introduces VitalLens, an app that estimates vital signs such as heart rate and respiration rate from selfie video in real time. VitalLens uses a computer vision model trained on a diverse dataset of video and physiological sensor data. We benchmark performance on several diverse datasets, including VV-Medium, which consists of 289 unique participants. VitalLens outperforms several existing methods including POS and MTTS-CAN on all datasets while maintaining a fast inference speed. On VV-Medium, VitalLens achieves absolute errors of 0.71 bpm for heart rate estimation, and 0.76 rpm for respiratory rate estimation.
Authors: Pareesa Ameneh Golnari
Diffusion models, such as Stable Diffusion (SD), offer the ability to generate high-resolution images with diverse features, but they come at a significant computational and memory cost. In classifier-free guided diffusion models, prolonged inference times are attributed to the necessity of computing two separate diffusion models at each denoising step. Recent work has shown promise in improving inference time through distillation techniques, teaching the model to perform similar denoising steps with reduced computations. However, the application of distillation introduces additional memory overhead to these already resource-intensive diffusion models, making it less practical.
To address these challenges, our research explores a novel approach that combines Low-Rank Adaptation (LoRA) with model distillation to efficiently compress diffusion models. This approach not only reduces inference time but also mitigates memory overhead, and notably decreases memory consumption even before applying distillation. The results are remarkable, featuring a significant reduction in inference time due to the distillation process and a substantial 50% reduction in memory consumption. Our examination of the generated images underscores that the incorporation of LoRA-enhanced distillation maintains image quality and alignment with the provided prompts. In summary, while conventional distillation tends to increase memory consumption, LoRA-enhanced distillation offers optimization without any trade-offs or compromises in quality.
Authors: Gourav Datta, Zeyu Liu, James Diffenderfer, Bhavya Kailkhura, Peter A. Beerel
Bio-inspired Spiking Neural Networks (SNN) are now demonstrating comparable accuracy to intricate convolutional neural networks (CNN), all while delivering remarkable energy and latency efficiency when deployed on neuromorphic hardware. In particular, ANN-to-SNN conversion has recently gained significant traction in developing deep SNNs with close to state-of-the-art (SOTA) test accuracy on complex image recognition tasks. However, advanced ANN-to-SNN conversion approaches demonstrate that for lossless conversion, the number of SNN time steps must equal the number of quantization steps in the ANN activation function. Reducing the number of time steps significantly increases the conversion error. Moreover, the spiking activity of the SNN, which dominates the compute energy in neuromorphic chips, does not reduce proportionally with the number of time steps. To mitigate the accuracy concern, we propose a novel ANN-to-SNN conversion framework, that incurs an exponentially lower number of time steps compared to that required in the SOTA conversion approaches. Our framework modifies the SNN integrate-and-fire (IF) neuron model with identical complexity and shifts the bias term of each batch normalization (BN) layer in the trained ANN. To mitigate the spiking activity concern, we propose training the source ANN with a fine-grained L1 regularizer with surrogate gradients that encourages high spike sparsity in the converted SNN. Our proposed framework thus yields lossless SNNs with ultra-low latency, ultra-low compute energy, thanks to the ultra-low timesteps and high spike sparsity, and ultra-high test accuracy, for example, 73.30% with only 4 time steps on the ImageNet dataset.
Authors: Soham Bafana, Radha Raghuraman, S. Abid Hussaini
Understanding object recognition patterns in mice is crucial for advancing behavioral neuroscience and has significant implications for human health, particularly in the realm of Alzheimer's research. This study is centered on the development, application, and evaluation of a state-of-the-art computational pipeline designed to analyze such behaviors, specifically focusing on Novel Object Recognition (NOR) and Spontaneous Location Recognition (SLR) tasks. The pipeline integrates three advanced computational models: Any-Maze for initial data collection, DeepLabCut for detailed pose estimation, and Convolutional Neural Networks (CNNs) for nuanced behavioral classification. Employed across four distinct mouse groups, this pipeline demonstrated high levels of accuracy and robustness. Despite certain challenges like video quality limitations and the need for manual calculations, the results affirm the pipeline's efficacy and potential for scalability. The study serves as a proof of concept for a multidimensional computational approach to behavioral neuroscience, emphasizing the pipeline's versatility and readiness for future, more complex analyses.
Authors: Issam Serraoui, Eric Granger, Abdenour Hadid, Abdelmalik Taleb-Ahmed
Automatic pain intensity estimation plays a pivotal role in healthcare and medical fields. While many methods have been developed to gauge human pain using behavioral or physiological indicators, facial expressions have emerged as a prominent tool for this purpose. Nevertheless, the dependence on labeled data for these techniques often renders them expensive and time-consuming. To tackle this, we introduce the Adaptive Hierarchical Spatio-temporal Dynamic Image (AHDI) technique. AHDI encodes spatiotemporal changes in facial videos into a singular RGB image, permitting the application of simpler 2D deep models for video representation. Within this framework, we employ a residual network to derive generalized facial representations. These representations are optimized for two tasks: estimating pain intensity and differentiating between genuine and simulated pain expressions. For the former, a regression model is trained using the extracted representations, while for the latter, a binary classifier identifies genuine versus feigned pain displays. Testing our method on two widely-used pain datasets, we observed encouraging results for both tasks. On the UNBC database, we achieved an MSE of 0.27 outperforming the SOTA which had an MSE of 0.40. On the BioVid dataset, our model achieved an accuracy of 89.76%, which is an improvement of 5.37% over the SOTA accuracy. Most notably, for distinguishing genuine from simulated pain, our accuracy stands at 94.03%, marking a substantial improvement of 8.98%. Our methodology not only minimizes the need for extensive labeled data but also augments the precision of pain evaluations, facilitating superior pain management.
Authors: Fatemeh Dehghani, Loutfouz Zaman
Emotion detection is a crucial component of Games User Research (GUR), as it allows game developers to gain insights into players' emotional experiences and tailor their games accordingly. However, detecting emotions in Virtual Reality (VR) games is challenging due to the Head-Mounted Display (HMD) that covers the top part of the player's face, namely, their eyes and eyebrows, which provide crucial information for recognizing the impression. To tackle this we used a Convolutional Neural Network (CNN) to train a model to predict emotions in full-face images where the eyes and eyebrows are covered. We used the FER2013 dataset, which we modified to cover eyes and eyebrows in images. The model in these images can accurately recognize seven different emotions which are anger, happiness, disgust, fear, impartiality, sadness and surprise.
We assessed the model's performance by testing it on two VR games and using it to detect players' emotions. We collected self-reported emotion data from the players after the gameplay sessions. We analyzed the data collected from our experiment to understand which emotions players experience during the gameplay. We found that our approach has the potential to enhance gameplay analysis by enabling the detection of players' emotions in VR games, which can help game developers create more engaging and immersive game experiences.
Authors: Dongliang Luo, Yuliang Liu, Rui Yang, Xianjin Liu, Jishen Zeng, Yu Zhou, Xiang Bai
With the surge in realistic text tampering, detecting fraudulent text in images has gained prominence for maintaining information security. However, the high costs associated with professional text manipulation and annotation limit the availability of real-world datasets, with most relying on synthetic tampering, which inadequately replicates real-world tampering attributes. To address this issue, we present the Real Text Manipulation (RTM) dataset, encompassing 14,250 text images, which include 5,986 manually and 5,258 automatically tampered images, created using a variety of techniques, alongside 3,006 unaltered text images for evaluating solution stability. Our evaluations indicate that existing methods falter in text forgery detection on the RTM dataset. We propose a robust baseline solution featuring a Consistency-aware Aggregation Hub and a Gated Cross Neighborhood-attention Fusion module for efficient multi-modal information fusion, supplemented by a Tampered-Authentic Contrastive Learning module during training, enriching feature representation distinction. This framework, extendable to other dual-stream architectures, demonstrated notable localization performance improvements of 7.33% and 6.38% on manual and overall manipulations, respectively. Our contributions aim to propel advancements in real-world text tampering detection. Code and dataset will be made available at https://github.com/DrLuo/RTM
Authors: Jacob Fein-Ashley, Tian Ye, Rajgopal Kannan, Viktor Prasanna, Carl Busart
Synthetic Aperture Radar SAR Automatic Target Recognition ATR is a key technique of remote-sensing image recognition which can be supported by deep neural networks The existing works of SAR ATR mostly focus on improving the accuracy of the target recognition while ignoring the systems performance in terms of speed and storage which is critical to real-world applications of SAR ATR For decision-makers aiming to identify a proper deep learning model to deploy in a SAR ATR system it is important to understand the performance of different candidate deep learning models and determine the best model accordingly This paper comprehensively benchmarks several advanced deep learning models for SAR ATR with multiple distinct SAR imagery datasets Specifically we train and test five SAR image classifiers based on Residual Neural Networks ResNet18 ResNet34 ResNet50 Graph Neural Network GNN and Vision Transformer for Small-Sized Datasets (SS-ViT) We select three datasets MSTAR GBSAR and SynthWakeSAR that offer heterogeneity We evaluate and compare the five classifiers concerning their classification accuracy runtime performance in terms of inference throughput and analytical performance in terms of number of parameters number of layers model size and number of operations Experimental results show that the GNN classifier outperforms with respect to throughput and latency However it is also shown that no clear model winner emerges from all of our chosen metrics and a one model rules all case is doubtful in the domain of SAR ATR
Authors: Jingchun Zhou, Tianyu Liang, Zongxin He, Dehuan Zhang, Weishi Zhang, Xianping Fu, Chongyi Li
Neural Radiance Field (NeRF) technology demonstrates immense potential in novel viewpoint synthesis tasks, due to its physics-based volumetric rendering process, which is particularly promising in underwater scenes. Addressing the limitations of existing underwater NeRF methods in handling light attenuation caused by the water medium and the lack of real Ground Truth (GT) supervision, this study proposes WaterHE-NeRF. We develop a new water-ray tracing field by Retinex theory that precisely encodes color, density, and illuminance attenuation in three-dimensional space. WaterHE-NeRF, through its illuminance attenuation mechanism, generates both degraded and clear multi-view images and optimizes image restoration by combining reconstruction loss with Wasserstein distance. Additionally, the use of histogram equalization (HE) as pseudo-GT enhances the network's accuracy in preserving original details and color distribution. Extensive experiments on real underwater datasets and synthetic datasets validate the effectiveness of WaterHE-NeRF. Our code will be made publicly available.
Authors: Kangneng Zhou, Daiheng Gao, Xuan Wang, Jie Zhang, Peng Zhang, Xusen Sun, Longhao Zhang, Shiqi Yang, Bang Zhang, Liefeng Bo, Yaxing Wang
Recently, 3D-aware face editing has witnessed remarkable progress. Although current approaches successfully perform mask-guided or text-based editing, these properties have not been combined into a single method. To address this limitation, we propose \textbf{MaTe3D}: mask-guided text-based 3D-aware portrait editing. First, we propose a new SDF-based 3D generator. To better perform masked-based editing (mainly happening in local areas), we propose SDF and density consistency losses, aiming to effectively model both the global and local representations jointly. Second, we introduce an inference-optimized method. We introduce two techniques based on the SDS (Score Distillation Sampling), including a blending SDS and a conditional SDS. The former aims to overcome the mismatch problem between geometry and appearance, ultimately harming fidelity. The conditional SDS contributes to further producing satisfactory and stable results. Additionally, we create CatMask-HQ dataset, a large-scale high-resolution cat face annotations. We perform experiments on both the FFHQ and CatMask-HQ datasets to demonstrate the effectiveness of the proposed method. Our method generates faithfully a edited 3D-aware face image given a modified mask and a text prompt. Our code and models will be publicly released.
Authors: Thong Nguyen, Xiaobao Wu, Xinshuai Dong, Khoi Le, Zhiyuan Hu, Cong-Duy Nguyen, See-Kiong Ng, Luu Anh Tuan
Fully fine-tuning pretrained large-scale transformer models has become a popular paradigm for video-language modeling tasks, such as temporal language grounding and video-language summarization. With a growing number of tasks and limited training data, such full fine-tuning approach leads to costly model storage and unstable training. To overcome these shortcomings, we introduce lightweight adapters to the pre-trained model and only update them at fine-tuning time. However, existing adapters fail to capture intrinsic temporal relations among video frames or textual words. Moreover, they neglect the preservation of critical task-related information that flows from the raw video-language input into the adapter's low-dimensional space. To address these issues, we first propose a novel REcurrent ADapter (READ) that employs recurrent computation to enable temporal modeling capability. Second, we propose Partial Video-Language Alignment (PVLA) objective via the use of partial optimal transport to maintain task-related information flowing into our READ modules. We validate our READ-PVLA framework through extensive experiments where READ-PVLA significantly outperforms all existing fine-tuning strategies on multiple low-resource temporal language grounding and video-language summarization benchmarks.
Authors: Jingchun Zhou, Qilin Gai, Weishi Zhang, Kin-man Lam, Xianping Fu, Ting Li, Chongyi Li
In underwater environments, variations in suspended particle concentration and turbidity cause severe image degradation, posing significant challenges to image enhancement (IE) and object detection (OD) tasks. Currently, in-air image enhancement and detection methods have made notable progress, but their application in underwater conditions is limited due to the complexity and variability of these environments. Fine-tuning in-air models saves high overhead and has more optional reference work than building an underwater model from scratch. To address these issues, we design a transfer plugin with multiple priors for converting in-air models to underwater applications, named IA2U. IA2U enables efficient application in underwater scenarios, thereby improving performance in Underwater IE and OD. IA2U integrates three types of underwater priors: the water type prior that characterizes the degree of image degradation, such as color and visibility; the degradation prior, focusing on differences in details and textures; and the sample prior, considering the environmental conditions at the time of capture and the characteristics of the photographed object. Utilizing a Transformer-like structure, IA2U employs these priors as query conditions and a joint task loss function to achieve hierarchical enhancement of task-level underwater image features, therefore considering the requirements of two different tasks, IE and OD. Experimental results show that IA2U combined with an in-air model can achieve superior performance in underwater image enhancement and object detection tasks. The code will be made publicly available.
Authors: Henrik Skibbe, Michal Byra, Akiya Watakabe, Tetsuo Yamamori, Marco Reisert
We introduce "PatchMorph," an new stochastic deep learning algorithm tailored for unsupervised 3D brain image registration. Unlike other methods, our method uses compact patches of a constant small size to derive solutions that can combine global transformations with local deformations. This approach minimizes the memory footprint of the GPU during training, but also enables us to operate on numerous amounts of randomly overlapping small patches during inference to mitigate image and patch boundary problems. PatchMorph adeptly handles world coordinate transformations between two input images, accommodating variances in attributes such as spacing, array sizes, and orientations. The spatial resolution of patches transitions from coarse to fine, addressing both global and local attributes essential for aligning the images. Each patch offers a unique perspective, together converging towards a comprehensive solution. Experiments on human T1 MRI brain images and marmoset brain images from serial 2-photon tomography affirm PatchMorph's superior performance.
Authors: Utkarsh Mall, Cheng Perng Phoo, Meilin Kelsey Liu, Carl Vondrick, Bharath Hariharan, Kavita Bala
We introduce a method to train vision-language models for remote-sensing images without using any textual annotations. Our key insight is to use co-located internet imagery taken on the ground as an intermediary for connecting remote-sensing images and language. Specifically, we train an image encoder for remote sensing images to align with the image encoder of CLIP using a large amount of paired internet and satellite images. Our unsupervised approach enables the training of a first-of-its-kind large-scale vision language model (VLM) for remote sensing images at two different resolutions. We show that these VLMs enable zero-shot, open-vocabulary image classification, retrieval, segmentation and visual question answering for satellite images. On each of these tasks, our VLM trained without textual annotations outperforms existing VLMs trained with supervision, with gains of up to 20% for classification and 80% for segmentation.
Authors: Chaoya Jiang, Haiyang Xu, Mengfan Dong, Jiaxing Chen, Wei Ye, Ming Yan, Qinghao Ye, Ji Zhang, Fei Huang, Shikun Zhang
Multi-modal large language models (MLLMs) have been shown to efficiently integrate natural language with visual information to handle multi-modal tasks. However, MLLMs still face a fundamental limitation of hallucinations, where they tend to generate erroneous or fabricated information. In this paper, we address hallucinations in MLLMs from a novel perspective of representation learning. We first analyzed the representation distribution of textual and visual tokens in MLLM, revealing two important findings: 1) there is a significant gap between textual and visual representations, indicating unsatisfactory cross-modal representation alignment; 2) representations of texts that contain and do not contain hallucinations are entangled, making it challenging to distinguish them. These two observations inspire us with a simple yet effective method to mitigate hallucinations. Specifically, we introduce contrastive learning into MLLMs and use text with hallucination as hard negative examples, naturally bringing representations of non-hallucinative text and visual samples closer while pushing way representations of non-hallucinating and hallucinative text. We evaluate our method quantitatively and qualitatively, showing its effectiveness in reducing hallucination occurrences and improving performance across multiple benchmarks. On the MMhal-Bench benchmark, our method obtains a 34.66% /29.5% improvement over the baseline MiniGPT-4/LLaVA.
Authors: Jie Xiao, Kai Zhu, Han Zhang, Zhiheng Liu, Yujun Shen, Yu Liu, Xueyang Fu, Zheng-Jun Zha
Consistency Models (CMs) have showed a promise in creating visual content efficiently and with high quality. However, the way to add new conditional controls to the pretrained CMs has not been explored. In this technical report, we consider alternative strategies for adding ControlNet-like conditional control to CMs and present three significant findings. 1) ControlNet trained for diffusion models (DMs) can be directly applied to CMs for high-level semantic controls but struggles with low-level detail and realism control. 2) CMs serve as an independent class of generative models, based on which ControlNet can be trained from scratch using Consistency Training proposed by Song et al. 3) A lightweight adapter can be jointly optimized under multiple conditions through Consistency Training, allowing for the swift transfer of DMs-based ControlNet to CMs. We study these three solutions across various conditional controls, including edge, depth, human pose, low-resolution image and masked image with text-to-image latent consistency models.
Authors: Bodong Zhang, Hamid Manoochehri, Man Minh Ho, Fahimeh Fooladgar, Yosep Chong, Beatrice S. Knudsen, Deepika Sirohi, Tolga Tasdizen
Histopathological image classification is one of the critical aspects in medical image analysis. Due to the high expense associated with the labeled data in model training, semi-supervised learning methods have been proposed to alleviate the need of extensively labeled datasets. In this work, we propose a model for semi-supervised classification tasks on digital histopathological Hematoxylin and Eosin (H&E) images. We call the new model Contrastive Learning with Adaptive Stain Separation and MixUp (CLASSMix). Our model is formed by two main parts: contrastive learning between adaptively stain separated Hematoxylin images and Eosin images, and pseudo labeling using MixUp. We compare our model with other state-of-the-art models on clear cell renal cell carcinoma (ccRCC) datasets from our institution and The Cancer Genome Atlas Program (TCGA). We demonstrate that our CLASSMix model has the best performance on both datasets. The contributions of different parts in our model are also analyzed.
Authors: Gauri Bhardwaj, Yuvaraj Govindarajulu, Sundaraparipurnan Narayanan, Pavan Kulkarni, Manojkumar Parmar
Medical imaging has revolutionized disease diagnosis, yet the potential is hampered by limited access to diverse and privacy-conscious datasets. Open-source medical datasets, while valuable, suffer from data quality and clinical information disparities. Generative models, such as diffusion models, aim to mitigate these challenges. At Stanford, researchers explored the utility of a fine-tuned Stable Diffusion model (RoentGen) for medical imaging data augmentation. Our work examines specific considerations to expand the Stanford research question, Could Stable Diffusion Solve a Gap in Medical Imaging Data? from the lens of bias and validity of the generated outcomes. We leveraged RoentGen to produce synthetic Chest-XRay (CXR) images and conducted assessments on bias, validity, and hallucinations. Diagnostic accuracy was evaluated by a disease classifier, while a COVID classifier uncovered latent hallucinations. The bias analysis unveiled disparities in classification performance among various subgroups, with a pronounced impact on the Female Hispanic subgroup. Furthermore, incorporating race and gender into input prompts exacerbated fairness issues in the generated images. The quality of synthetic images exhibited variability, particularly in certain disease classes, where there was more significant uncertainty compared to the original images. Additionally, we observed latent hallucinations, with approximately 42% of the images incorrectly indicating COVID, hinting at the presence of hallucinatory elements. These identifications provide new research directions towards interpretability of synthetic CXR images, for further understanding of associated risks and patient safety in medical applications.
Authors: Guangfeng Jiang, Jun Liu, Yuzhi Wu, Wenlong Liao, Tao He, Pai Peng
Instance segmentation is a fundamental research in computer vision, especially in autonomous driving. However, manual mask annotation for instance segmentation is quite time-consuming and costly. To address this problem, some prior works attempt to apply weakly supervised manner by exploring 2D or 3D boxes. However, no one has ever successfully segmented 2D and 3D instances simultaneously by only using 2D box annotations, which could further reduce the annotation cost by an order of magnitude. Thus, we propose a novel framework called Multimodal Weakly Supervised Instance Segmentation (MWSIS), which incorporates various fine-grained label generation and correction modules for both 2D and 3D modalities to improve the quality of pseudo labels, along with a new multimodal cross-supervision approach, named Consistency Sparse Cross-modal Supervision (CSCS), to reduce the inconsistency of multimodal predictions by response distillation. Particularly, transferring the 3D backbone to downstream tasks not only improves the performance of the 3D detectors, but also outperforms fully supervised instance segmentation with only 5% fully supervised annotations. On the Waymo dataset, the proposed framework demonstrates significant improvements over the baseline, especially achieving 2.59% mAP and 12.75% mAP increases for 2D and 3D instance segmentation tasks, respectively. The code is available at https://github.com/jiangxb98/mwsis-plugin.
Authors: Jonathan J.Y. Kim, Martin Urschler, Patricia J. Riddle, Jorg S. Wicker
With the advancement in robotics, it is becoming increasingly common for large factories and warehouses to incorporate visual SLAM (vSLAM) enabled automated robots that operate closely next to humans. This makes any adversarial attacks on vSLAM components potentially detrimental to humans working alongside them. Loop Closure Detection (LCD) is a crucial component in vSLAM that minimizes the accumulation of drift in mapping, since even a small drift can accumulate into a significant drift over time. A prior work by Kim et al., SymbioLCD2, unified visual features and semantic objects into a single graph structure for finding loop closure candidates. While this provided a performance improvement over visual feature-based LCD, it also created a single point of vulnerability for potential graph-based adversarial attacks. Unlike previously reported visual-patch based attacks, small graph perturbations are far more challenging to detect, making them a more significant threat. In this paper, we present Adversarial-LCD, a novel black-box evasion attack framework that employs an eigencentrality-based perturbation method and an SVM-RBF surrogate model with a Weisfeiler-Lehman feature extractor for attacking graph-based LCD. Our evaluation shows that the attack performance of Adversarial-LCD with the SVM-RBF surrogate model was superior to that of other machine learning surrogate algorithms, including SVM-linear, SVM-polynomial, and Bayesian classifier, demonstrating the effectiveness of our attack framework. Furthermore, we show that our eigencentrality-based perturbation method outperforms other algorithms, such as Random-walk and Shortest-path, highlighting the efficiency of Adversarial-LCD's perturbation selection method.
Authors: Jinsong Shi, Pan Gao, Jie Qin
Image Quality Assessment (IQA) has long been a research hotspot in the field of image processing, especially No-Reference Image Quality Assessment (NR-IQA). Due to the powerful feature extraction ability, existing Convolution Neural Network (CNN) and Transformers based NR-IQA methods have achieved considerable progress. However, they still exhibit limited capability when facing unknown authentic distortion datasets. To further improve NR-IQA performance, in this paper, a novel supervised contrastive learning (SCL) and Transformer-based NR-IQA model SaTQA is proposed. We first train a model on a large-scale synthetic dataset by SCL (no image subjective score is required) to extract degradation features of images with various distortion types and levels. To further extract distortion information from images, we propose a backbone network incorporating the Multi-Stream Block (MSB) by combining the CNN inductive bias and Transformer long-term dependence modeling capability. Finally, we propose the Patch Attention Block (PAB) to obtain the final distorted image quality score by fusing the degradation features learned from contrastive learning with the perceptual distortion information extracted by the backbone network. Experimental results on seven standard IQA datasets show that SaTQA outperforms the state-of-the-art methods for both synthetic and authentic datasets. Code is available at https://github.com/I2-Multimedia-Lab/SaTQA
Authors: Jingchun Zhou, Zongxin He, Dehuan Zhang, Kin-man Lam, Weishi Zhang, Xianping Fu, Yi Wang, Chongyi Li
Underwater image enhancement (UIE) is a challenging task due to the complex degradation caused by underwater environments. To solve this issue, previous methods often idealize the degradation process, and neglect the impact of medium noise and object motion on the distribution of image features, limiting the generalization and adaptability of the model. Previous methods use the reference gradient that is constructed from original images and synthetic ground-truth images. This may cause the network performance to be influenced by some low-quality training data. Our approach utilizes predicted images to dynamically update pseudo-labels, adding a dynamic gradient to optimize the network's gradient space. This process improves image quality and avoids local optima. Moreover, we propose a Feature Restoration and Reconstruction module (FRR) based on a Channel Combination Inference (CCI) strategy and a Frequency Domain Smoothing module (FRS). These modules decouple other degradation features while reducing the impact of various types of noise on network performance. Experiments on multiple public datasets demonstrate the superiority of our method over existing state-of-the-art approaches, especially in achieving performance milestones: PSNR of 25.6dB and SSIM of 0.93 on the UIEB dataset. Its efficiency in terms of parameter size and inference time further attests to its broad practicality. The code will be made publicly available.
Authors: Zeming Chen, Wenwei Zhang, Xinjiang Wang, Kai Chen, Zhi Wang
While the pseudo-label method has demonstrated considerable success in semi-supervised object detection tasks, this paper uncovers notable limitations within this approach. Specifically, the pseudo-label method tends to amplify the inherent strengths of the detector while accentuating its weaknesses, which is manifested in the missed detection of pseudo-labels, particularly for small and tail category objects. To overcome these challenges, this paper proposes Mixed Pseudo Labels (MixPL), consisting of Mixup and Mosaic for pseudo-labeled data, to mitigate the negative impact of missed detections and balance the model's learning across different object scales. Additionally, the model's detection performance on tail categories is improved by resampling labeled data with relevant instances. Notably, MixPL consistently improves the performance of various detectors and obtains new state-of-the-art results with Faster R-CNN, FCOS, and DINO on COCO-Standard and COCO-Full benchmarks. Furthermore, MixPL also exhibits good scalability on large models, improving DINO Swin-L by 2.5% mAP and achieving nontrivial new records (60.2% mAP) on the COCO val2017 benchmark without extra annotations.
Authors: Kongming Liang, Xinran Wang, Rui Wang, Donghui Gao, Ling Jin, Weidong Liu, Xiatian Zhu, Zhanyu Ma, Jun Guo
Attribute labeling at large scale is typically incomplete and partial, posing significant challenges to model optimization. Existing attribute learning methods often treat the missing labels as negative or simply ignore them all during training, either of which could hamper the model performance to a great extent. To overcome these limitations, in this paper we leverage the available vision-language knowledge to explicitly disclose the missing labels for enhancing model learning. Given an image, we predict the likelihood of each missing attribute label assisted by an off-the-shelf vision-language model, and randomly select to ignore those with high scores in training. Our strategy strikes a good balance between fully ignoring and negatifying the missing labels, as these high scores are found to be informative on revealing label ambiguity. Extensive experiments show that our proposed vision-language assisted loss can achieve state-of-the-art performance on the newly cleaned VAW dataset. Qualitative evaluation demonstrates the ability of the proposed method in predicting more complete attributes.
Authors: Yuwei Guo, Wenhao Zhang, Licheng Jiao, Shuang Wang, Shuo Wang, Fang Liu
Visible-infrared person re-identification (VI-ReID) aims to search the same pedestrian of interest across visible and infrared modalities. Existing models mainly focus on compensating for modality-specific information to reduce modality variation. However, these methods often lead to a higher computational overhead and may introduce interfering information when generating the corresponding images or features. To address this issue, it is critical to leverage pedestrian-attentive features and learn modality-complete and -consistent representation. In this paper, a novel Transferring Modality-Aware Pedestrian Attentive Learning (TMPA) model is proposed, focusing on the pedestrian regions to efficiently compensate for missing modality-specific features. Specifically, we propose a region-based data augmentation module PedMix to enhance pedestrian region coherence by mixing the corresponding regions from different modalities. A lightweight hybrid compensation module, i.e., the Modality Feature Transfer (MFT), is devised to integrate cross attention and convolution networks to fully explore the discriminative modality-complete features with minimal computational overhead. Extensive experiments conducted on the benchmark SYSU-MM01 and RegDB datasets demonstrated the effectiveness of our proposed TMPA model.
Authors: Weiguang Zhao, Guanyu Yang, Chaolong Yang, Chenru Jiang, Yuyao Yan, Rui Zhang, Kaizhu Huang
With the explosive 3D data growth, the urgency of utilizing zero-shot learning to facilitate data labeling becomes evident. Recently, the methods via transferring Contrastive Language-Image Pre-training (CLIP) to 3D vision have made great progress in the 3D zero-shot classification task. However, these methods primarily focus on aligned pose 3D objects (ap-3os), overlooking the recognition of 3D objects with open poses (op-3os) typically encountered in real-world scenarios, such as an overturned chair or a lying teddy bear. To this end, we propose a more challenging benchmark for 3D open-pose zero-shot classification. Echoing our benchmark, we design a concise angle-refinement mechanism that automatically optimizes one ideal pose as well as classifies these op-3os. Furthermore, we make a first attempt to bridge 2D pre-trained diffusion model as a classifer to 3D zero-shot classification without any additional training. Such 2D diffusion to 3D objects proves vital in improving zero-shot classification for both ap-3os and op-3os. Our model notably improves by 3.5% and 15.8% on ModelNet10$^{\ddag}$ and McGill$^{\ddag}$ open pose benchmarks, respectively, and surpasses the current state-of-the-art by 6.8% on the aligned pose ModelNet10, affirming diffusion's efficacy in 3D zero-shot tasks.
Authors: Yuke Zhu, Yumeng Ruan, Zihua Xiong, Sheng Guo
Regression loss design is an essential topic for oriented object detection. Due to the periodicity of the angle and the ambiguity of width and height definition, traditional L1-distance loss and its variants have been suffered from the metric discontinuity and the square-like problem. As a solution, the distribution based methods show significant advantages by representing oriented boxes as distributions. Differing from exploited the Gaussian distribution to get analytical form of distance measure, we propose a novel oriented regression loss, Wasserstein Distance(EWD) loss, to alleviate the square-like problem. Specifically, for the oriented box(OBox) representation, we choose a specially-designed distribution whose probability density function is only nonzero over the edges. On this basis, we develop Wasserstein distance as the measure. Besides, based on the edge representation of OBox, the EWD loss can be generalized to quadrilateral and polynomial regression scenarios. Experiments on multiple popular datasets and different detectors show the effectiveness of the proposed method.
Authors: Yuchen Yang, Yu Qiao, Xiao Sun
Automatic estimation of 3D human pose from monocular RGB images is a challenging and unsolved problem in computer vision. In a supervised manner, approaches heavily rely on laborious annotations and present hampered generalization ability due to the limited diversity of 3D pose datasets. To address these challenges, we propose a unified framework that leverages mask as supervision for unsupervised 3D pose estimation. With general unsupervised segmentation algorithms, the proposed model employs skeleton and physique representations that exploit accurate pose information from coarse to fine. Compared with previous unsupervised approaches, we organize the human skeleton in a fully unsupervised way which enables the processing of annotation-free data and provides ready-to-use estimation results. Comprehensive experiments demonstrate our state-of-the-art pose estimation performance on Human3.6M and MPI-INF-3DHP datasets. Further experiments on in-the-wild datasets also illustrate the capability to access more data to boost our model. Code will be available at https://github.com/Charrrrrlie/Mask-as-Supervision.
Authors: Xiao Ma, Zetian Zhang, Zexuan Ji, Kun Huang, Na Su, Songtao Yuan, Qiang Chen
Myopia is a manifestation of visual impairment caused by an excessively elongated eyeball. Image data is critical material for studying high myopia and pathological myopia. Measurements of spherical equivalent and axial length are the gold standards for identifying high myopia, but the available image data for matching them is scarce. In addition, the criteria for defining high myopia vary from study to study, and therefore the inclusion of samples in automated screening efforts requires an appropriate assessment of interpretability. In this work, we propose a model called adjustable robust transformer (ARTran) for high myopia screening of optical coherence tomography (OCT) data. Based on vision transformer, we propose anisotropic patch embedding (APE) to capture more discriminative features of high myopia. To make the model effective under variable screening conditions, we propose an adjustable class embedding (ACE) to replace the fixed class token, which changes the output to adapt to different conditions. Considering the confusion of the data at high myopia and low myopia threshold, we introduce the label noise learning strategy and propose a shifted subspace transition matrix (SST) to enhance the robustness of the model. Besides, combining the two structures proposed above, the model can provide evidence for uncertainty evaluation. The experimental results demonstrate the effectiveness and reliability of the proposed method. Code is available at: https://github.com/maxiao0234/ARTran.
Authors: Jingyang Xiang, Siqi Li, Junhao Chen, Zhuangzhi Chen, Tianxin Huang, Linpeng Peng, Yong Liu
N:M sparsity has received increasing attention due to its remarkable performance and latency trade-off compared with structured and unstructured sparsity. However, existing N:M sparsity methods do not differentiate the relative importance of weights among blocks and leave important weights underappreciated. Besides, they directly apply N:M sparsity to the whole network, which will cause severe information loss. Thus, they are still sub-optimal. In this paper, we propose an efficient and effective Multi-Axis Query methodology, dubbed as MaxQ, to rectify these problems. During the training, MaxQ employs a dynamic approach to generate soft N:M masks, considering the weight importance across multiple axes. This method enhances the weights with more importance and ensures more effective updates. Meanwhile, a sparsity strategy that gradually increases the percentage of N:M weight blocks is applied, which allows the network to heal from the pruning-induced damage progressively. During the runtime, the N:M soft masks can be precomputed as constants and folded into weights without causing any distortion to the sparse pattern and incurring additional computational overhead. Comprehensive experiments demonstrate that MaxQ achieves consistent improvements across diverse CNN architectures in various computer vision tasks, including image classification, object detection and instance segmentation. For ResNet50 with 1:16 sparse pattern, MaxQ can achieve 74.6\% top-1 accuracy on ImageNet and improve by over 2.8\% over the state-of-the-art.
Authors: Guanxing Lu, Ziwei Wang, Changliu Liu, Jiwen Lu, Yansong Tang
Embodied Instruction Following (EIF) requires agents to complete human instruction by interacting objects in complicated surrounding environments. Conventional methods directly consider the sparse human instruction to generate action plans for agents, which usually fail to achieve human goals because of the instruction incoherence in action descriptions. On the contrary, we propose ThinkBot that reasons the thought chain in human instruction to recover the missing action descriptions, so that the agent can successfully complete human goals by following the coherent instruction. Specifically, we first design an instruction completer based on large language models to recover the missing actions with interacted objects between consecutive human instruction, where the perceived surrounding environments and the completed sub-goals are considered for instruction completion. Based on the partially observed scene semantic maps, we present an object localizer to infer the position of interacted objects for agents to achieve complex human goals. Extensive experiments in the simulated environment show that our ThinkBot outperforms the state-of-the-art EIF methods by a sizable margin in both success rate and execution efficiency.
Authors: Xianghui Xie, Bharat Lal Bhatnagar, Jan Eric Lenssen, Gerard Pons-Moll
Reconstructing human-object interaction in 3D from a single RGB image is a challenging task and existing data driven methods do not generalize beyond the objects present in the carefully curated 3D interaction datasets. Capturing large-scale real data to learn strong interaction and 3D shape priors is very expensive due to the combinatorial nature of human-object interactions. In this paper, we propose ProciGen (Procedural interaction Generation), a method to procedurally generate datasets with both, plausible interaction and diverse object variation. We generate 1M+ human-object interaction pairs in 3D and leverage this large-scale data to train our HDM (Hierarchical Diffusion Model), a novel method to reconstruct interacting human and unseen objects, without any templates. Our HDM is an image-conditioned diffusion model that learns both realistic interaction and highly accurate human and object shapes. Experiments show that our HDM trained with ProciGen significantly outperforms prior methods that requires template meshes and that our dataset allows training methods with strong generalization ability to unseen object instances. Our code and data will be publicly released at: https://virtualhumans.mpi-inf.mpg.de/procigen-hdm.
Authors: Manuel Röder, Leon Heller, Maximilian Münch, Frank-Michael Schleif
With the advent of interconnected and sensor-equipped edge devices, Federated Learning (FL) has gained significant attention, enabling decentralized learning while maintaining data privacy. However, FL faces two challenges in real-world tasks: expensive data labeling and domain shift between source and target samples. In this paper, we introduce a privacy-preserving, resource-efficient FL concept for client adaptation in hardware-constrained environments. Our approach includes server model pre-training on source data and subsequent fine-tuning on target data via low-end clients. The local client adaptation process is streamlined by probabilistic mixing of instance-level feature statistics approximated from source and target domain data. The adapted parameters are transferred back to the central server and globally aggregated. Preliminary results indicate that our method reduces computational and transmission costs while maintaining competitive performance on downstream tasks.
Authors: Shengguang Wu, Mei Yuan, Qi Su
Recent advances in image and video creation, especially AI-based image synthesis, have led to the production of numerous visual scenes that exhibit a high level of abstractness and diversity. Consequently, Visual Storytelling (VST), a task that involves generating meaningful and coherent narratives from a collection of images, has become even more challenging and is increasingly desired beyond real-world imagery. While existing VST techniques, which typically use autoregressive decoders, have made significant progress, they suffer from low inference speed and are not well-suited for synthetic scenes. To this end, we propose a novel diffusion-based system DiffuVST, which models the generation of a series of visual descriptions as a single conditional denoising process. The stochastic and non-autoregressive nature of DiffuVST at inference time allows it to generate highly diverse narratives more efficiently. In addition, DiffuVST features a unique design with bi-directional text history guidance and multimodal adapter modules, which effectively improve inter-sentence coherence and image-to-text fidelity. Extensive experiments on the story generation task covering four fictional visual-story datasets demonstrate the superiority of DiffuVST over traditional autoregressive models in terms of both text quality and inference speed.
Authors: Qian Li, Yuxiao Hu, Yinpeng Dong, Dongxiao Zhang, Yuntian Chen
Adversarial training is often formulated as a min-max problem, however, concentrating only on the worst adversarial examples causes alternating repetitive confusion of the model, i.e., previously defended or correctly classified samples are not defensible or accurately classifiable in subsequent adversarial training. We characterize such non-ignorable samples as "hiders", which reveal the hidden high-risk regions within the secure area obtained through adversarial training and prevent the model from finding the real worst cases. We demand the model to prevent hiders when defending against adversarial examples for improving accuracy and robustness simultaneously. By rethinking and redefining the min-max optimization problem for adversarial training, we propose a generalized adversarial training algorithm called Hider-Focused Adversarial Training (HFAT). HFAT introduces the iterative evolution optimization strategy to simplify the optimization problem and employs an auxiliary model to reveal hiders, effectively combining the optimization directions of standard adversarial training and prevention hiders. Furthermore, we introduce an adaptive weighting mechanism that facilitates the model in adaptively adjusting its focus between adversarial examples and hiders during different training periods. We demonstrate the effectiveness of our method based on extensive experiments, and ensure that HFAT can provide higher robustness and accuracy.
Authors: Ziqiang Zhang, Yan Yan, Jing-Hao Xue, Hanzi Wang
Most existing GAN inversion methods either achieve accurate reconstruction but lack editability or offer strong editability at the cost of fidelity. Hence, how to balance the distortioneditability trade-off is a significant challenge for GAN inversion. To address this challenge, we introduce a novel spatial-contextual discrepancy information compensationbased GAN-inversion method (SDIC), which consists of a discrepancy information prediction network (DIPN) and a discrepancy information compensation network (DICN). SDIC follows a "compensate-and-edit" paradigm and successfully bridges the gap in image details between the original image and the reconstructed/edited image. On the one hand, DIPN encodes the multi-level spatial-contextual information of the original and initial reconstructed images and then predicts a spatial-contextual guided discrepancy map with two hourglass modules. In this way, a reliable discrepancy map that models the contextual relationship and captures finegrained image details is learned. On the other hand, DICN incorporates the predicted discrepancy information into both the latent code and the GAN generator with different transformations, generating high-quality reconstructed/edited images. This effectively compensates for the loss of image details during GAN inversion. Both quantitative and qualitative experiments demonstrate that our proposed method achieves the excellent distortion-editability trade-off at a fast inference speed for both image inversion and editing tasks.
Authors: Yi Sun, Xin Xu, Jian Li, Guanglei Xie, Yifei Shi, Qiang Fang
It's challenging to balance the networks stability and plasticity in continual learning scenarios, considering stability suffers from the update of model and plasticity benefits from it. Existing works usually focus more on the stability and restrict the learning plasticity of later tasks to avoid catastrophic forgetting of learned knowledge. Differently, we propose a continual learning method named Split2MetaFusion which can achieve better trade-off by employing a two-stage strategy: splitting and meta-weighted fusion. In this strategy, a slow model with better stability, and a fast model with better plasticity are learned sequentially at the splitting stage. Then stability and plasticity are both kept by fusing the two models in an adaptive manner. Towards this end, we design an optimizer named Task-Preferred Null Space Projector(TPNSP) to the slow learning process for narrowing the fusion gap. To achieve better model fusion, we further design a Dreaming-Meta-Weighted fusion policy for better maintaining the old and new knowledge simultaneously, which doesn't require to use the previous datasets. Experimental results and analysis reported in this work demonstrate the superiority of the proposed method for maintaining networks stability and keeping its plasticity. Our code will be released.
Authors: Peng Liu, Fanyi Wang, Jingwen Su, Yanhao Zhang, Guojun Qi
Existing saliency object detection (SOD) methods struggle to satisfy fast inference and accurate results simultaneously in high resolution scenes. They are limited by the quality of public datasets and efficient network modules for high-resolution images. To alleviate these issues, we propose to construct a saliency object matting dataset HRSOM and a lightweight network PSUNet. Considering efficient inference of mobile depolyment framework, we design a symmetric pixel shuffle module and a lightweight module TRSU. Compared to 13 SOD methods, the proposed PSUNet has the best objective performance on the high-resolution benchmark dataset. Evaluation results of objective assessment are superior compared to U$^2$Net that has 10 times of parameter amount of our network. On Snapdragon 8 Gen 2 Mobile Platform, inference a single 640$\times$640 image only takes 113ms. And on the subjective assessment, evaluation results are better than the industry benchmark IOS16 (Lift subject from background).
Authors: Kaipeng Zheng, Weiran Huang, Lichao Sun
Few-shot learning has been studied to adapt models to tasks with very few samples. It holds profound significance, particularly in clinical tasks, due to the high annotation cost of medical images. Several works have explored few-shot learning on medical images, yet they still require a large number of medical images for pre-training models to gain domain-specific priors. Vision foundation models recently have achieved remarkable success in natural images. Hence, adapting rapidly advancing vision foundation models from natural images to few-shot clinical tasks holds great promise. MedFMC has recently organized a challenge to shed more light on this topic at NeurIPS 2023. In this work, we present our challenge solution. We observe that a simple variant of fine-tuning with partial freezing shows remarkable performance. Empirical evidence demonstrates that this approach could outperform various common fine-tuning methods under limited sample sizes. Additionally, we explore enhanced utilization of semantic supervision to boost performance. We propose a novel approach that contextualizes labels via large language models (LLMs). Our findings reveal that the context generated by LLMs significantly enhances the discrimination of semantic embeddings for similar categories, resulting in a notable performance improvement of 3%-5% in 1-shot settings compared to commonly employed one-hot labels and other semantic supervision methods. Our solution secures the 1st place in the MedFMC challenge.
Authors: Jing Xu
Although transformer is preferred in natural language processing, few studies have applied it in the field of medical imaging. For its long-term dependency, the transformer is expected to contribute to unconventional convolution neural net conquer their inherent spatial induction bias. The lately suggested transformer-based partition method only uses the transformer as an auxiliary module to help encode the global context into a convolutional representation. There is hardly any study about how to optimum bond self-attention (the kernel of transformers) with convolution. To solve the problem, the article proposes MS-Twins (Multi-Scale Twins), which is a powerful segmentation model on account of the bond of self-attention and convolution. MS-Twins can better capture semantic and fine-grained information by combining different scales and cascading features. Compared with the existing network structure, MS-Twins has made significant progress on the previous method based on the transformer of two in common use data sets, Synapse and ACDC. In particular, the performance of MS-Twins on Synapse is 8% higher than SwinUNet. Even compared with nnUNet, the best entirely convoluted medical image segmentation network, the performance of MS-Twins on Synapse and ACDC still has a bit advantage.
Authors: Xiaochuan Li, Baoyu Fan, Runze Zhang, Liang Jin, Di Wang, Zhenhua Guo, Yaqian Zhao, Rengang Li
The emergence of ChatGPT has once again sparked research in generative artificial intelligence (GAI). While people have been amazed by the generated results, they have also noticed the reasoning potential reflected in the generated textual content. However, this current ability for causal reasoning is primarily limited to the domain of language generation, such as in models like GPT-3. In visual modality, there is currently no equivalent research. Considering causal reasoning in visual content generation is significant. This is because visual information contains infinite granularity. Particularly, images can provide more intuitive and specific demonstrations for certain reasoning tasks, especially when compared to coarse-grained text. Hence, we propose a new image generation task called visual question answering with image (VQAI) and establish a dataset of the same name based on the classic \textit{Tom and Jerry} animated series. Additionally, we develop a new paradigm for image generation to tackle the challenges of this task. Finally, we perform extensive experiments and analyses, including visualizations of the generated content and discussions on the potentials and limitations. The code and data are publicly available under the license of CC BY-NC-SA 4.0 for academic and non-commercial usage. The code and dataset are publicly available at: https://github.com/IEIT-AGI/MIX-Shannon/blob/main/projects/VQAI/lgd_vqai.md.
Authors: Abdelrahman Eldesokey, Peter Wonka
We propose a zero-shot approach for consistent Text-to-Animated-Characters synthesis based on pre-trained Text-to-Image (T2I) diffusion models. Existing Text-to-Video (T2V) methods are expensive to train and require large-scale video datasets to produce diverse characters and motions. At the same time, their zero-shot alternatives fail to produce temporally consistent videos. We strive to bridge this gap, and we introduce a zero-shot approach that produces temporally consistent videos of animated characters and requires no training or fine-tuning. We leverage existing text-based motion diffusion models to generate diverse motions that we utilize to guide a T2I model. To achieve temporal consistency, we introduce the Spatial Latent Alignment module that exploits cross-frame dense correspondences that we compute to align the latents of the video frames. Furthermore, we propose Pixel-Wise Guidance to steer the diffusion process in a direction that minimizes visual discrepancies. Our proposed approach generates temporally consistent videos with diverse motions and styles, outperforming existing zero-shot T2V approaches in terms of pixel-wise consistency and user preference.
Authors: I-Jieh Liu, Ci-Siang Lin, Fu-En Yang, Yu-Chiang Frank Wang
Federated Learning (FL) is an emerging paradigm that enables multiple users to collaboratively train a robust model in a privacy-preserving manner without sharing their private data. Most existing approaches of FL only consider traditional single-label image classification, ignoring the impact when transferring the task to multi-label image classification. Nevertheless, it is still challenging for FL to deal with user heterogeneity in their local data distribution in the real-world FL scenario, and this issue becomes even more severe in multi-label image classification. Inspired by the recent success of Transformers in centralized settings, we propose a novel FL framework for multi-label classification. Since partial label correlation may be observed by local clients during training, direct aggregation of locally updated models would not produce satisfactory performances. Thus, we propose a novel FL framework of Language-Guided Transformer (FedLGT) to tackle this challenging task, which aims to exploit and transfer knowledge across different clients for learning a robust global model. Through extensive experiments on various multi-label datasets (e.g., FLAIR, MS-COCO, etc.), we show that our FedLGT is able to achieve satisfactory performance and outperforms standard FL techniques under multi-label FL scenarios. Code is available at https://github.com/Jack24658735/FedLGT.
Authors: Aayush Singh, Aayush J Rana, Akash Kumar, Shruti Vyas, Yogesh Singh Rawat
In this work, we focus on label efficient learning for video action detection. We develop a novel semi-supervised active learning approach which utilizes both labeled as well as unlabeled data along with informative sample selection for action detection. Video action detection requires spatio-temporal localization along with classification, which poses several challenges for both active learning informative sample selection as well as semi-supervised learning pseudo label generation. First, we propose NoiseAug, a simple augmentation strategy which effectively selects informative samples for video action detection. Next, we propose fft-attention, a novel technique based on high-pass filtering which enables effective utilization of pseudo label for SSL in video action detection by emphasizing on relevant activity region within a video. We evaluate the proposed approach on three different benchmark datasets, UCF-101-24, JHMDB-21, and Youtube-VOS. First, we demonstrate its effectiveness on video action detection where the proposed approach outperforms prior works in semi-supervised and weakly-supervised learning along with several baseline approaches in both UCF101-24 and JHMDB-21. Next, we also show its effectiveness on Youtube-VOS for video object segmentation demonstrating its generalization capability for other dense prediction tasks in videos.
Authors: Ri Cheng, Ruian He, Xuhao Jiang, Shili Zhou, Weimin Tan, Bo Yan
Existing recurrent optical flow estimation networks are computationally expensive since they use a fixed large number of iterations to update the flow field for each sample. An efficient network should skip iterations when the flow improvement is limited. In this paper, we develop a Context-Aware Iteration Policy Network for efficient optical flow estimation, which determines the optimal number of iterations per sample. The policy network achieves this by learning contextual information to realize whether flow improvement is bottlenecked or minimal. On the one hand, we use iteration embedding and historical hidden cell, which include previous iterations information, to convey how flow has changed from previous iterations. On the other hand, we use the incremental loss to make the policy network implicitly perceive the magnitude of optical flow improvement in the subsequent iteration. Furthermore, the computational complexity in our dynamic network is controllable, allowing us to satisfy various resource preferences with a single trained model. Our policy network can be easily integrated into state-of-the-art optical flow networks. Extensive experiments show that our method maintains performance while reducing FLOPs by about 40%/20% for the Sintel/KITTI datasets.
Authors: Yuda Zou, Xin Xiao, Peilin Zhou, Zhichao Sun, Bo Du, Yongchao Xu
Object counting is a field of growing importance in domains such as security surveillance, urban planning, and biology. The annotation is usually provided in terms of 2D points. However, the complexity of object shapes and subjective of annotators may lead to annotation inconsistency, potentially confusing the model during training. To alleviate this issue, we introduce the Noised Autoencoders (NAE) methodology, which extracts general positional knowledge from all annotations. The method involves adding random offsets to initial point annotations, followed by a UNet to restore them to their original positions. Similar to MAE, NAE faces challenges in restoring non-generic points, necessitating reliance on the most common positions inferred from general knowledge. This reliance forms the cornerstone of our method's effectiveness. Different from existing noise-resistance methods, our approach focus on directly improving initial point annotations. Extensive experiments show that NAE yields more consistent annotations compared to the original ones, steadily enhancing the performance of advanced models trained with these revised annotations. \textbf{Remarkably, the proposed approach helps to set new records in nine datasets}. We will make the NAE codes and refined point annotations available.
Authors: Ilektra Karasante, Lazaro Alonso, Ioannis Prapas, Akanksha Ahuja, Nuno Carvalhais, Ioannis Papoutsis
The global occurrence, scale, and frequency of wildfires pose significant threats to ecosystem services and human livelihoods. To effectively quantify and attribute the antecedent conditions for wildfires, a thorough understanding of Earth system dynamics is imperative. In response, we introduce the SeasFire datacube, a meticulously curated spatiotemporal dataset tailored for global sub-seasonal to seasonal wildfire modeling via Earth observation. The SeasFire datacube comprises of 59 variables encompassing climate, vegetation, oceanic indices, and human factors, has an 8-day temporal resolution and a spatial resolution of 0.25 degrees, and spans from 2001 to 2021. We showcase the versatility of SeasFire for exploring the variability and seasonality of wildfire drivers, modeling causal links between ocean-climate teleconnections and wildfires, and predicting sub-seasonal wildfire patterns across multiple timescales with a Deep Learning model. We publicly release the SeasFire datacube and appeal to Earth system scientists and Machine Learning practitioners to use it for an improved understanding and anticipation of wildfires.
Authors: Xiaojie Fang, Xingguo Song, Xiangyin Meng, Xu Fang, Sheng Jin
The low-level spatial detail information and high-level semantic abstract information are both essential to the semantic segmentation task. The features extracted by the deep network can obtain rich semantic information, while a lot of spatial information is lost. However, how to recover spatial detail information effectively and fuse it with high-level semantics has not been well addressed so far. In this paper, we propose a new architecture based on Bilateral Segmentation Network (BiseNet) called Multi-scale Covariance Feature Fusion Network (MCFNet). Specifically, this network introduces a new feature refinement module and a new feature fusion module. Furthermore, a gating unit named L-Gate is proposed to filter out invalid information and fuse multi-scale features. We evaluate our proposed model on Cityscapes, CamVid datasets and compare it with the state-of-the-art methods. Extensive experiments show that our method achieves competitive success. On Cityscapes, we achieve 75.5% mIOU with a speed of 151.3 FPS.
Authors: Yuanbin Wang, Shaofei Huang, Yulu Gao, Zhen Wang, Rui Wang, Kehua Sheng, Bo Zhang, Si Liu
Traditional 3D segmentation methods can only recognize a fixed range of classes that appear in the training set, which limits their application in real-world scenarios due to the lack of generalization ability. Large-scale visual-language pre-trained models, such as CLIP, have shown their generalization ability in the zero-shot 2D vision tasks, but are still unable to be applied to 3D semantic segmentation directly. In this work, we focus on zero-shot point cloud semantic segmentation and propose a simple yet effective baseline to transfer the visual-linguistic knowledge implied in CLIP to point cloud encoder at both feature and output levels. Both feature-level and output-level alignments are conducted between 2D and 3D encoders for effective knowledge transfer. Concretely, a Multi-granularity Cross-modal Feature Alignment (MCFA) module is proposed to align 2D and 3D features from global semantic and local position perspectives for feature-level alignment. For the output level, per-pixel pseudo labels of unseen classes are extracted using the pre-trained CLIP model as supervision for the 3D segmentation model to mimic the behavior of the CLIP image encoder. Extensive experiments are conducted on two popular benchmarks of point cloud segmentation. Our method outperforms significantly previous state-of-the-art methods under zero-shot setting (+29.2% mIoU on SemanticKITTI and 31.8% mIoU on nuScenes), and further achieves promising results in the annotation-free point cloud semantic segmentation setting, showing its great potential for label-efficient learning.
Authors: Kai Pan, Linyang Li, Li Lin, Pujin Cheng, Junyan Lyu, Lei Xi, Xiaoyin Tang
Photoacoustic Microscopy (PAM) images integrating the advantages of optical contrast and acoustic resolution have been widely used in brain studies. However, there exists a trade-off between scanning speed and image resolution. Compared with traditional raster scanning, rotational scanning provides good opportunities for fast PAM imaging by optimizing the scanning mechanism. Recently, there is a trend to incorporate deep learning into the scanning process to further increase the scanning speed.Yet, most such attempts are performed for raster scanning while those for rotational scanning are relatively rare. In this study, we propose a novel and well-performing super-resolution framework for rotational scanning-based PAM imaging. To eliminate adjacent rows' displacements due to subject motion or high-frequency scanning distortion,we introduce a registration module across odd and even rows in the preprocessing and incorporate displacement degradation in the training. Besides, gradient-based patch selection is proposed to increase the probability of blood vessel patches being selected for training. A Transformer-based network with a global receptive field is applied for better performance. Experimental results on both synthetic and real datasets demonstrate the effectiveness and generalizability of our proposed framework for rotationally scanned PAM images'super-resolution, both quantitatively and qualitatively. Code is available at https://github.com/11710615/PAMSR.git.
Authors: Shentong Mo, Enze Xie, Yue Wu, Junsong Chen, Matthias Nießner, Zhenguo Li
Diffusion Transformers have recently shown remarkable effectiveness in generating high-quality 3D point clouds. However, training voxel-based diffusion models for high-resolution 3D voxels remains prohibitively expensive due to the cubic complexity of attention operators, which arises from the additional dimension of voxels. Motivated by the inherent redundancy of 3D compared to 2D, we propose FastDiT-3D, a novel masked diffusion transformer tailored for efficient 3D point cloud generation, which greatly reduces training costs. Specifically, we draw inspiration from masked autoencoders to dynamically operate the denoising process on masked voxelized point clouds. We also propose a novel voxel-aware masking strategy to adaptively aggregate background/foreground information from voxelized point clouds. Our method achieves state-of-the-art performance with an extreme masking ratio of nearly 99%. Moreover, to improve multi-category 3D generation, we introduce Mixture-of-Expert (MoE) in 3D diffusion model. Each category can learn a distinct diffusion path with different experts, relieving gradient conflict. Experimental results on the ShapeNet dataset demonstrate that our method achieves state-of-the-art high-fidelity and diverse 3D point cloud generation performance. Our FastDiT-3D improves 1-Nearest Neighbor Accuracy and Coverage metrics when generating 128-resolution voxel point clouds, using only 6.5% of the original training cost.
Authors: Renyang Liu, Wei Zhou, Xin Jin, Song Gao, Yuanyu Wang, Ruxin Wang
In generating adversarial examples, the conventional black-box attack methods rely on sufficient feedback from the to-be-attacked models by repeatedly querying until the attack is successful, which usually results in thousands of trials during an attack. This may be unacceptable in real applications since Machine Learning as a Service Platform (MLaaS) usually only returns the final result (i.e., hard-label) to the client and a system equipped with certain defense mechanisms could easily detect malicious queries. By contrast, a feasible way is a hard-label attack that simulates an attacked action being permitted to conduct a limited number of queries. To implement this idea, in this paper, we bypass the dependency on the to-be-attacked model and benefit from the characteristics of the distributions of adversarial examples to reformulate the attack problem in a distribution transform manner and propose a distribution transform-based attack (DTA). DTA builds a statistical mapping from the benign example to its adversarial counterparts by tackling the conditional likelihood under the hard-label black-box settings. In this way, it is no longer necessary to query the target model frequently. A well-trained DTA model can directly and efficiently generate a batch of adversarial examples for a certain input, which can be used to attack un-seen models based on the assumed transferability. Furthermore, we surprisingly find that the well-trained DTA model is not sensitive to the semantic spaces of the training dataset, meaning that the model yields acceptable attack performance on other datasets. Extensive experiments validate the effectiveness of the proposed idea and the superiority of DTA over the state-of-the-art.
Authors: Sunghwan Hong, Jaewoo Jung, Heeseong Shin, Jiaolong Yang, Seungryong Kim, Chong Luo
This work delves into the task of pose-free novel view synthesis from stereo pairs, a challenging and pioneering task in 3D vision. Our innovative framework, unlike any before, seamlessly integrates 2D correspondence matching, camera pose estimation, and NeRF rendering, fostering a synergistic enhancement of these tasks. We achieve this through designing an architecture that utilizes a shared representation, which serves as a foundation for enhanced 3D geometry understanding. Capitalizing on the inherent interplay between the tasks, our unified framework is trained end-to-end with the proposed training strategy to improve overall model accuracy. Through extensive evaluations across diverse indoor and outdoor scenes from two real-world datasets, we demonstrate that our approach achieves substantial improvement over previous methodologies, especially in scenarios characterized by extreme viewpoint changes and the absence of accurate camera poses.
Authors: Jiacheng Ruan, Jingsheng Gao, Mingye Xie, Suncheng Xiang, Zefang Yu, Ting Liu, Yuzhuo Fu
The Parameter-Efficient Fine-Tuning (PEFT) method, which adjusts or introduces fewer trainable parameters to calibrate pre-trained models on downstream tasks, has become a recent research interest. However, existing PEFT methods within the traditional fine-tiuning framework have two main shortcomings: 1) They overlook the explicit association between trainable parameters and downstream task knowledge. 2) They neglect the interaction between the intrinsic task-agnostic knowledge of pre-trained models and the task-specific knowledge in downstream tasks. To address this gap, we propose a novel fine-tuning framework, named GIST, in a plug-and-play manner. Specifically, our framework first introduces a trainable token, called the Gist token, when applying PEFT methods on downstream tasks. This token serves as an aggregator of the task-specific knowledge learned by the PEFT methods and forms an explicit association with downstream knowledge. Furthermore, to facilitate explicit interaction between task-agnostic and task-specific knowledge, we introduce the concept of Knowledge Interaction via a Bidirectional Kullback-Leibler Divergence objective. As a result, PEFT methods within our framework can make the pre-trained model understand downstream tasks more comprehensively by leveraging the knowledge interaction. Extensive experiments demonstrate the universality and scalability of our framework. Notably, on the VTAB-1K benchmark, we employ the Adapter (a prevalent PEFT method) within our GIST framework and achieve a performance boost of 2.25%, with an increase of only 0.8K parameters. The Code will be released.
Authors: Renyang Liu, Wei Zhou, Sixin Wu, Jun Zhao, Kwok-Yan Lam
Extensive studies have demonstrated that deep neural networks (DNNs) are vulnerable to adversarial attacks, which brings a huge security risk to the further application of DNNs, especially for the AI models developed in the real world. Despite the significant progress that has been made recently, existing attack methods still suffer from the unsatisfactory performance of escaping from being detected by naked human eyes due to the formulation of adversarial example (AE) heavily relying on a noise-adding manner. Such mentioned challenges will significantly increase the risk of exposure and result in an attack to be failed. Therefore, in this paper, we propose the Salient Spatially Transformed Attack (SSTA), a novel framework to craft imperceptible AEs, which enhance the stealthiness of AEs by estimating a smooth spatial transform metric on a most critical area to generate AEs instead of adding external noise to the whole image. Compared to state-of-the-art baselines, extensive experiments indicated that SSTA could effectively improve the imperceptibility of the AEs while maintaining a 100\% attack success rate.
Authors: Yuliang Gu, Zhichao Sun, Xin Xiao, Yuda Zou, Zelong Liu, Yongchao Xu
Semi-supervised image segmentation has attracted great attention recently. The key is how to leverage unlabeled images in the training process. Most methods maintain consistent predictions of the unlabeled images under variations (e.g., adding noise/perturbations, or creating alternative versions) in the image and/or model level. In most image-level variation, medical images often have prior structure information, which has not been well explored. In this paper, we propose novel dual structure-preserving image filterings (DSPIF) as the image-level variations for semi-supervised medical image segmentation. Motivated by connected filtering that simplifies image via filtering in structure-aware tree-based image representation, we resort to the dual contrast invariant Max-tree and Min-tree representation. Specifically, we propose a novel connected filtering that removes topologically equivalent nodes (i.e. connected components) having no siblings in the Max/Min-tree. This results in two filtered images preserving topologically critical structure. Applying such dual structure-preserving image filterings in mutual supervision is beneficial for semi-supervised medical image segmentation. Extensive experimental results on three benchmark datasets demonstrate that the proposed method significantly/consistently outperforms some state-of-the-art methods. The source codes will be publicly available.
Authors: Joonhyun Jeong, Geondo Park, Jayeon Yoo, Hyungsik Jung, Heesu Kim
Open-vocabulary object detection (OVOD) aims to recognize novel objects whose categories are not included in training set. In order to classify these unseen classes during training, many OVOD frameworks leverage the zero-shot capability of largely pretrained vision and language models, such as CLIP. To further improve generalization on the unseen novel classes, several approaches proposed to additionally train with pseudo region labeling on the external data sources that contain a substantial number of novel category labels beyond the existing training data. Albeit its simplicity, these pseudo-labeling methods still exhibit limited improvement with regard to the genuine novel classes that were not pseudo-labeled. In this paper, we present a novel, yet simple technique that helps generalization on the overall distribution of novel classes. Inspired by our observation that numerous novel classes reside within the convex hull constructed by the base (seen) classes in the CLIP embedding space, we propose to synthesize proxy-novel classes approximating novel classes via linear mixup between a pair of base classes. By training our detector with these synthetic proxy-novel classes, we effectively explore the embedding space of novel classes. The experimental results on various OVOD benchmarks such as LVIS and COCO demonstrate superior performance on novel classes compared to the other state-of-the-art methods.
Authors: Tuan Truong, Farnaz Khun Jush, Matthias Lenga
Near- and duplicate image detection is a critical concern in the field of medical imaging. Medical datasets often contain similar or duplicate images from various sources, which can lead to significant performance issues and evaluation biases, especially in machine learning tasks due to data leakage between training and testing subsets. In this paper, we present an approach for identifying near- and duplicate 3D medical images leveraging publicly available 2D computer vision embeddings. We assessed our approach by comparing embeddings extracted from two state-of-the-art self-supervised pretrained models and two different vector index structures for similarity retrieval. We generate an experimental benchmark based on the publicly available Medical Segmentation Decathlon dataset. The proposed method yields promising results for near- and duplicate image detection achieving a mean sensitivity and specificity of 0.9645 and 0.8559, respectively.
Authors: Wenjie Yin, Yi Yu, Hang Yin, Danica Kragic, Mårten Björkman
Current training of motion style transfer systems relies on consistency losses across style domains to preserve contents, hindering its scalable application to a large number of domains and private data. Recent image transfer works show the potential of independent training on each domain by leveraging implicit bridging between diffusion models, with the content preservation, however, limited to simple data patterns. We address this by imposing biased sampling in backward diffusion while maintaining the domain independence in the training stage. We construct the bias from the source domain keyframes and apply them as the gradient of content constraints, yielding a framework with keyframe manifold constraint gradients (KMCGs). Our validation demonstrates the success of training separate models to transfer between as many as ten dance motion styles. Comprehensive experiments find a significant improvement in preserving motion contents in comparison to baseline and ablative diffusion-based style transfer models. In addition, we perform a human study for a subjective assessment of the quality of generated dance motions. The results validate the competitiveness of KMCGs.
Authors: Yoonwoo Jeong, Jinwoo Lee, Chiheon Kim, Minsu Cho, Doyup Lee
Transfer learning of large-scale Text-to-Image (T2I) models has recently shown impressive potential for Novel View Synthesis (NVS) of diverse objects from a single image. While previous methods typically train large models on multi-view datasets for NVS, fine-tuning the whole parameters of T2I models not only demands a high cost but also reduces the generalization capacity of T2I models in generating diverse images in a new domain. In this study, we propose an effective method, dubbed NVS-Adapter, which is a plug-and-play module for a T2I model, to synthesize novel multi-views of visual objects while fully exploiting the generalization capacity of T2I models. NVS-Adapter consists of two main components; view-consistency cross-attention learns the visual correspondences to align the local details of view features, and global semantic conditioning aligns the semantic structure of generated views with the reference view. Experimental results demonstrate that the NVS-Adapter can effectively synthesize geometrically consistent multi-views and also achieve high performance on benchmarks without full fine-tuning of T2I models. The code and data are publicly available in ~\href{https://postech-cvlab.github.io/nvsadapter/}{https://postech-cvlab.github.io/nvsadapter/}.
Authors: Tomáš Souček, Dima Damen, Michael Wray, Ivan Laptev, Josef Sivic
We address the task of generating temporally consistent and physically plausible images of actions and object state transformations. Given an input image and a text prompt describing the targeted transformation, our generated images preserve the environment and transform objects in the initial image. Our contributions are threefold. First, we leverage a large body of instructional videos and automatically mine a dataset of triplets of consecutive frames corresponding to initial object states, actions, and resulting object transformations. Second, equipped with this data, we develop and train a conditioned diffusion model dubbed GenHowTo. Third, we evaluate GenHowTo on a variety of objects and actions and show superior performance compared to existing methods. In particular, we introduce a quantitative evaluation where GenHowTo achieves 88% and 74% on seen and unseen interaction categories, respectively, outperforming prior work by a large margin.
Authors: Jian Zhu, Yu Cui, Zhangmin Huang, Xingyu Li, Lei Liu, Lingfang Zeng, Li-Rong Dai
The multi-view hash method converts heterogeneous data from multiple views into binary hash codes, which is one of the critical technologies in multimedia retrieval. However, the current methods mainly explore the complementarity among multiple views while lacking confidence learning and fusion. Moreover, in practical application scenarios, the single-view data contain redundant noise. To conduct the confidence learning and eliminate unnecessary noise, we propose a novel Adaptive Confidence Multi-View Hashing (ACMVH) method. First, a confidence network is developed to extract useful information from various single-view features and remove noise information. Furthermore, an adaptive confidence multi-view network is employed to measure the confidence of each view and then fuse multi-view features through a weighted summation. Lastly, a dilation network is designed to further enhance the feature representation of the fused features. To the best of our knowledge, we pioneer the application of confidence learning into the field of multimedia retrieval. Extensive experiments on two public datasets show that the proposed ACMVH performs better than state-of-the-art methods (maximum increase of 3.24%). The source code is available at https://github.com/HackerHyper/ACMVH.
Authors: Alexandros Graikos, Srikar Yellapragada, Minh-Quan Le, Saarthak Kapse, Prateek Prasanna, Joel Saltz, Dimitris Samaras
To synthesize high-fidelity samples, diffusion models typically require auxiliary data to guide the generation process. However, it is impractical to procure the painstaking patch-level annotation effort required in specialized domains like histopathology and satellite imagery; it is often performed by domain experts and involves hundreds of millions of patches. Modern-day self-supervised learning (SSL) representations encode rich semantic and visual information. In this paper, we posit that such representations are expressive enough to act as proxies to fine-grained human labels. We introduce a novel approach that trains diffusion models conditioned on embeddings from SSL. Our diffusion models successfully project these features back to high-quality histopathology and remote sensing images. In addition, we construct larger images by assembling spatially consistent patches inferred from SSL embeddings, preserving long-range dependencies. Augmenting real data by generating variations of real images improves downstream classifier accuracy for patch-level and larger, image-scale classification tasks. Our models are effective even on datasets not encountered during training, demonstrating their robustness and generalizability. Generating images from learned embeddings is agnostic to the source of the embeddings. The SSL embeddings used to generate a large image can either be extracted from a reference image, or sampled from an auxiliary model conditioned on any related modality (e.g. class labels, text, genomic data). As proof of concept, we introduce the text-to-large image synthesis paradigm where we successfully synthesize large pathology and satellite images out of text descriptions.
Authors: Hansong Zhang, Shikun Li, Dan Zeng, Chenggang Yan, Shiming Ge
As the size of the datasets getting larger, accurately annotating such datasets is becoming more impractical due to the expensiveness on both time and economy. Therefore, crowd-sourcing has been widely adopted to alleviate the cost of collecting labels, which also inevitably introduces label noise and eventually degrades the performance of the model. To learn from crowd-sourcing annotations, modeling the expertise of each annotator is a common but challenging paradigm, because the annotations collected by crowd-sourcing are usually highly-sparse. To alleviate this problem, we propose Coupled Confusion Correction (CCC), where two models are simultaneously trained to correct the confusion matrices learned by each other. Via bi-level optimization, the confusion matrices learned by one model can be corrected by the distilled data from the other. Moreover, we cluster the ``annotator groups'' who share similar expertise so that their confusion matrices could be corrected together. In this way, the expertise of the annotators, especially of those who provide seldom labels, could be better captured. Remarkably, we point out that the annotation sparsity not only means the average number of labels is low, but also there are always some annotators who provide very few labels, which is neglected by previous works when constructing synthetic crowd-sourcing annotations. Based on that, we propose to use Beta distribution to control the generation of the crowd-sourcing labels so that the synthetic annotations could be more consistent with the real-world ones. Extensive experiments are conducted on two types of synthetic datasets and three real-world datasets, the results of which demonstrate that CCC significantly outperforms state-of-the-art approaches.
Authors: Jiyoung Kim, Kyuhong Shim, Insu Lee, Byonghyo Shim
Unsupervised semantic segmentation (USS) aims to discover and recognize meaningful categories without any labels. For a successful USS, two key abilities are required: 1) information compression and 2) clustering capability. Previous methods have relied on feature dimension reduction for information compression, however, this approach may hinder the process of clustering. In this paper, we propose a novel USS framework called Expand-and-Quantize Unsupervised Semantic Segmentation (EQUSS), which combines the benefits of high-dimensional spaces for better clustering and product quantization for effective information compression. Our extensive experiments demonstrate that EQUSS achieves state-of-the-art results on three standard benchmarks. In addition, we analyze the entropy of USS features, which is the first step towards understanding USS from the perspective of information theory.
Authors: Chinedu Innocent Nwoye, Kareem Elgohary, Anvita Srinivas, Fauzan Zaid, Joël L. Lavanchy, Nicolas Padoy
Tool tracking in surgical videos is vital in computer-assisted intervention for tasks like surgeon skill assessment, safety zone estimation, and human-machine collaboration during minimally invasive procedures. The lack of large-scale datasets hampers Artificial Intelligence implementation in this domain. Current datasets exhibit overly generic tracking formalization, often lacking surgical context: a deficiency that becomes evident when tools move out of the camera's scope, resulting in rigid trajectories that hinder realistic surgical representation. This paper addresses the need for a more precise and adaptable tracking formalization tailored to the intricacies of endoscopic procedures by introducing CholecTrack20, an extensive dataset meticulously annotated for multi-class multi-tool tracking across three perspectives representing the various ways of considering the temporal duration of a tool trajectory: (1) intraoperative, (2) intracorporeal, and (3) visibility within the camera's scope. The dataset comprises 20 laparoscopic videos with over 35,000 frames and 65,000 annotated tool instances with details on spatial location, category, identity, operator, phase, and surgical visual conditions. This detailed dataset caters to the evolving assistive requirements within a procedure.
Authors: Zihao Zhao, Yuxiao Liu, Han Wu, Yonghao Li, Sheng Wang, Lin Teng, Disheng Liu, Xiang Li, Zhiming Cui, Qian Wang, Dinggang Shen
Contrastive Language-Image Pre-training (CLIP), a straightforward yet effective pre-training paradigm, successfully introduces semantic-rich text supervision to vision models and has demonstrated promising results in various tasks due to its generalizability and interpretability. It has recently gained increasing interest in the medical imaging domain, either as a powerful pre-training paradigm for medical vision language alignment or a pre-trained key component for various clinical tasks. With the aim of facilitating a deeper understanding of this promising direction, this survey offers an in-depth exploration of the CLIP paradigm within the domain of medical imaging, regarding both refined CLIP pre-training and CLIP-driven applications. Our survey (1) starts with a brief introduction to the fundamentals of CLIP methodology. (2) Then, we investigate the adaptation of CLIP pre-training in the medical domain, focusing on how to optimize CLIP given characteristics of medical images and reports. (3) Furthermore, we explore the practical utilization of CLIP pre-trained models in various tasks, including classification, dense prediction, and cross-modal tasks. (4) Finally, we discuss existing limitations of CLIP in the context of medical imaging and propose forward-looking directions to address the demands of medical imaging domain. We expect that this comprehensive survey will provide researchers in the field of medical image analysis with a holistic understanding of the CLIP paradigm and its potential implications. The project page is available at https://github.com/zhaozh10/Awesome-CLIP-in-Medical-Imaging, which will be regularly updated.
Authors: Cyril Barrelet, Marc Chaumont, Gérard Subsol
In this paper we present a pipeline using stereo images in order to automatically identify, track in 3D fish, and measure fish population.
Authors: Johannes S. Fischer, Ming Gui, Pingchuan Ma, Nick Stracke, Stefan A. Baumann, Björn Ommer
Recently, there has been tremendous progress in visual synthesis and the underlying generative models. Here, diffusion models (DMs) stand out particularly, but lately, flow matching (FM) has also garnered considerable interest. While DMs excel in providing diverse images, they suffer from long training and slow generation. With latent diffusion, these issues are only partially alleviated. Conversely, FM offers faster training and inference but exhibits less diversity in synthesis. We demonstrate that introducing FM between the Diffusion model and the convolutional decoder offers high-resolution image synthesis with reduced computational cost and model size. Diffusion can then efficiently provide the necessary generation diversity. FM compensates for the lower resolution, mapping the small latent space to a high-dimensional one. Subsequently, the convolutional decoder of the LDM maps these latents to high-resolution images. By combining the diversity of DMs, the efficiency of FMs, and the effectiveness of convolutional decoders, we achieve state-of-the-art high-resolution image synthesis at $1024^2$ with minimal computational cost. Importantly, our approach is orthogonal to recent approximation and speed-up strategies for the underlying DMs, making it easily integrable into various DM frameworks.
Authors: Qiwei Tian, Chenhao Lin, Qian Li, Zhengyu Zhao, Chao Shen
Adversarial training has achieved substantial performance in defending image retrieval systems against adversarial examples. However, existing studies still suffer from two major limitations: model collapse and weak adversary. This paper addresses these two limitations by proposing collapse-oriented (COLO) adversarial training with triplet decoupling (TRIDE). Specifically, COLO prevents model collapse by temporally orienting the perturbation update direction with a new collapse metric, while TRIDE yields a strong adversary by spatially decoupling the update targets of perturbation into the anchor and the two candidates of a triplet. Experimental results demonstrate that our COLO-TRIDE outperforms the current state of the art by 7% on average over 10 robustness metrics and across 3 popular datasets. In addition, we identify the fairness limitations of commonly used robustness metrics in image retrieval and propose a new metric for more meaningful robustness evaluation. Codes will be made publicly available on GitHub.
Authors: Marwa Kechaou, Mokhtar Z. Alaya, Romain Hérault, Gilles Gasso
Adversarial learning baselines for domain adaptation (DA) approaches in the context of semantic segmentation are under explored in semi-supervised framework. These baselines involve solely the available labeled target samples in the supervision loss. In this work, we propose to enhance their usefulness on both semantic segmentation and the single domain classifier neural networks. We design new training objective losses for cases when labeled target data behave as source samples or as real target samples. The underlying rationale is that considering the set of labeled target samples as part of source domain helps reducing the domain discrepancy and, hence, improves the contribution of the adversarial loss. To support our approach, we consider a complementary method that mixes source and labeled target data, then applies the same adaptation process. We further propose an unsupervised selection procedure using entropy to optimize the choice of labeled target samples for adaptation. We illustrate our findings through extensive experiments on the benchmarks GTA5, SYNTHIA, and Cityscapes. The empirical evaluation highlights competitive performance of our proposed approach.
Authors: Jian Hu, Jiayi Lin, Weitong Cai, Shaogang Gong
Camouflaged object detection (COD) approaches heavily rely on pixel-level annotated datasets. Weakly-supervised COD (WSCOD) approaches use sparse annotations like scribbles or points to reduce annotation effort, but this can lead to decreased accuracy. The Segment Anything Model (SAM) shows remarkable segmentation ability with sparse prompts like points. However, manual prompt is not always feasible, as it may not be accessible in real-world application. Additionally, it only provides localization information instead of semantic one, which can intrinsically cause ambiguity in interpreting the targets. In this work, we aim to eliminate the need for manual prompt. The key idea is to employ Cross-modal Chains of Thought Prompting (CCTP) to reason visual prompts using the semantic information given by a generic text prompt.To that end, we introduce a test-time adaptation per-instance mechanism called Generalizable SAM (GenSAM) to automatically enerate and optimize visual prompts the generic task prompt for WSCOD. In particular, CCTP maps a single generic text prompt onto image-specific consensus foreground and background heatmaps using vision-language models, acquiring reliable visual prompts. Moreover, to test-time adapt the visual prompts, we further propose Progressive Mask Generation (PMG) to iteratively reweight the input image, guiding the model to focus on the targets in a coarse-to-fine manner. Crucially, all network parameters are fixed, avoiding the need for additional training. Experiments demonstrate the superiority of GenSAM. Experiments on three benchmarks demonstrate that GenSAM outperforms point supervision approaches and achieves comparable results to scribble supervision ones, solely relying on general task descriptions as prompts. our codes is in: https://lwpyh.github.io/GenSAM/.
Authors: Linglin Jing, Ying Xue, Xu Yan, Chaoda Zheng, Dong Wang, Ruimao Zhang, Zhigang Wang, Hui Fang, Bin Zhao, Zhen Li
The field of 4D point cloud understanding is rapidly developing with the goal of analyzing dynamic 3D point cloud sequences. However, it remains a challenging task due to the sparsity and lack of texture in point clouds. Moreover, the irregularity of point cloud poses a difficulty in aligning temporal information within video sequences. To address these issues, we propose a novel cross-modal knowledge transfer framework, called X4D-SceneFormer. This framework enhances 4D-Scene understanding by transferring texture priors from RGB sequences using a Transformer architecture with temporal relationship mining. Specifically, the framework is designed with a dual-branch architecture, consisting of an 4D point cloud transformer and a Gradient-aware Image Transformer (GIT). During training, we employ multiple knowledge transfer techniques, including temporal consistency losses and masked self-attention, to strengthen the knowledge transfer between modalities. This leads to enhanced performance during inference using single-modal 4D point cloud inputs. Extensive experiments demonstrate the superior performance of our framework on various 4D point cloud video understanding tasks, including action recognition, action segmentation and semantic segmentation. The results achieve 1st places, i.e., 85.3% (+7.9%) accuracy and 47.3% (+5.0%) mIoU for 4D action segmentation and semantic segmentation, on the HOI4D challenge\footnote{\url{this http URL}.}, outperforming previous state-of-the-art by a large margin. We release the code at https://github.com/jinglinglingling/X4D
Authors: Hallee E. Wong, Marianne Rakic, John Guttag, Adrian V. Dalca
Semantic medical image segmentation is a crucial part of both scientific research and clinical care. With enough labelled data, deep learning models can be trained to accurately automate specific medical image segmentation tasks. However, manually segmenting images to create training data is highly labor intensive. In this paper, we present ScribblePrompt, an interactive segmentation framework for medical imaging that enables human annotators to segment unseen structures using scribbles, clicks, and bounding boxes. Scribbles are an intuitive and effective form of user interaction for complex tasks, however most existing methods focus on click-based interactions. We introduce algorithms for simulating realistic scribbles that enable training models that are amenable to multiple types of interaction. To achieve generalization to new tasks, we train on a diverse collection of 65 open-access biomedical datasets -- using both real and synthetic labels. We test ScribblePrompt on multiple network architectures and unseen datasets, and demonstrate that it can be used in real-time on a single CPU. We evaluate ScribblePrompt using manually-collected scribbles, simulated interactions, and a user study. ScribblePrompt outperforms existing methods in all our evaluations. In the user study, ScribblePrompt reduced annotation time by 28% while improving Dice by 15% compared to existing methods. We showcase ScribblePrompt in an online demo and provide code at https://scribbleprompt.csail.mit.edu
Authors: Haoyu Tang, Han Jiang, Mingzhu Xu, Yupeng Hu, Jihua Zhu, Liqiang Nie
Recently, temporal action localization (TAL) has garnered significant interest in information retrieval community. However, existing supervised/weakly supervised methods are heavily dependent on extensive labeled temporal boundaries and action categories, which is labor-intensive and time-consuming. Although some unsupervised methods have utilized the ``iteratively clustering and localization'' paradigm for TAL, they still suffer from two pivotal impediments: 1) unsatisfactory video clustering confidence, and 2) unreliable video pseudolabels for model training. To address these limitations, we present a novel self-paced incremental learning model to enhance clustering and localization training simultaneously, thereby facilitating more effective unsupervised TAL. Concretely, we improve the clustering confidence through exploring the contextual feature-robust visual information. Thereafter, we design two (constant- and variable- speed) incremental instance learning strategies for easy-to-hard model training, thus ensuring the reliability of these video pseudolabels and further improving overall localization performance. Extensive experiments on two public datasets have substantiated the superiority of our model over several state-of-the-art competitors.
Authors: Haiming Zhang, Zhihao Yuan, Chaoda Zheng, Xu Yan, Baoyuan Wang, Guanbin Li, Song Wu, Shuguang Cui, Zhen Li
Although existing speech-driven talking face generation methods achieve significant progress, they are far from real-world application due to the avatar-specific training demand and unstable lip movements. To address the above issues, we propose the GSmoothFace, a novel two-stage generalized talking face generation model guided by a fine-grained 3d face model, which can synthesize smooth lip dynamics while preserving the speaker's identity. Our proposed GSmoothFace model mainly consists of the Audio to Expression Prediction (A2EP) module and the Target Adaptive Face Translation (TAFT) module. Specifically, we first develop the A2EP module to predict expression parameters synchronized with the driven speech. It uses a transformer to capture the long-term audio context and learns the parameters from the fine-grained 3D facial vertices, resulting in accurate and smooth lip-synchronization performance. Afterward, the well-designed TAFT module, empowered by Morphology Augmented Face Blending (MAFB), takes the predicted expression parameters and target video as inputs to modify the facial region of the target video without distorting the background content. The TAFT effectively exploits the identity appearance and background context in the target video, which makes it possible to generalize to different speakers without retraining. Both quantitative and qualitative experiments confirm the superiority of our method in terms of realism, lip synchronization, and visual quality. See the project page for code, data, and request pre-trained models: https://zhanghm1995.github.io/GSmoothFace.
Authors: Michael Lanier, Aayush Dhakal, Zhexiao Xiong, Arthur Li, Nathan Jacobs, Yevgeniy Vorobeychik
In critical operations where aerial imagery plays an essential role, the integrity and trustworthiness of data are paramount. The emergence of adversarial attacks, particularly those that exploit control over labels or employ physically feasible trojans, threatens to erode that trust, making the analysis and mitigation of these attacks a matter of urgency. We demonstrate how adversarial attacks can degrade confidence in geospatial systems, specifically focusing on scenarios where the attacker's control over labels is restricted and the use of realistic threat vectors. Proposing and evaluating several innovative attack methodologies, including those tailored to overhead images, we empirically show their threat to remote sensing systems using high-quality SpaceNet datasets. Our experimentation reflects the unique challenges posed by aerial imagery, and these preliminary results not only reveal the potential risks but also highlight the non-trivial nature of the problem compared to recent works.
Authors: Pinelopi Papalampidi, Skanda Koppula, Shreya Pathak, Justin Chiu, Joe Heyward, Viorica Patraucean, Jiajun Shen, Antoine Miech, Andrew Zisserman, Aida Nematzdeh
Understanding long, real-world videos requires modeling of long-range visual dependencies. To this end, we explore video-first architectures, building on the common paradigm of transferring large-scale, image--text models to video via shallow temporal fusion. However, we expose two limitations to the approach: (1) decreased spatial capabilities, likely due to poor video--language alignment in standard video datasets, and (2) higher memory consumption, bottlenecking the number of frames that can be processed. To mitigate the memory bottleneck, we systematically analyze the memory/accuracy trade-off of various efficient methods: factorized attention, parameter-efficient image-to-video adaptation, input masking, and multi-resolution patchification. Surprisingly, simply masking large portions of the video (up to 75%) during contrastive pre-training proves to be one of the most robust ways to scale encoders to videos up to 4.3 minutes at 1 FPS. Our simple approach for training long video-to-text models, which scales to 1B parameters, does not add new architectural complexity and is able to outperform the popular paradigm of using much larger LLMs as an information aggregator over segment-based information on benchmarks with long-range temporal dependencies (YouCook2, EgoSchema).
Authors: Chen Ju, Haicheng Wang, Zeqian Li, Xu Chen, Zhonghua Zhai, Weilin Huang, Shuai Xiao
Vision-Language Large Models (VLMs) have become primary backbone of AI, due to the impressive performance. However, their expensive computation costs, i.e., throughput and delay, impede potentials in real-world scenarios. To achieve acceleration for VLMs, most existing methods focus on the model perspective: pruning, distillation, quantification, but completely overlook the data-perspective redundancy. To fill the overlook, this paper pioneers the severity of data redundancy, and designs one plug-and-play Turbo module guided by information degree to prune inefficient tokens from visual or textual data. In pursuit of efficiency-performance trade-offs, information degree takes two key factors into consideration: mutual redundancy and semantic value. Concretely, the former evaluates the data duplication between sequential tokens; while the latter evaluates each token by its contribution to the overall semantics. As a result, tokens with high information degree carry less redundancy and stronger semantics. For VLMs' calculation, Turbo works as a user-friendly plug-in that sorts data referring to information degree, utilizing only top-level ones to save costs. Its advantages are multifaceted, e.g., being generally compatible to various VLMs across understanding and generation, simple use without retraining and trivial engineering efforts. On multiple public VLMs benchmarks, we conduct extensive experiments to reveal the gratifying acceleration of Turbo, under negligible performance drop.
Authors: Kaiwen Zhang, Yifan Zhou, Xudong Xu, Xingang Pan, Bo Dai
Diffusion models have achieved remarkable image generation quality surpassing previous generative models. However, a notable limitation of diffusion models, in comparison to GANs, is their difficulty in smoothly interpolating between two image samples, due to their highly unstructured latent space. Such a smooth interpolation is intriguing as it naturally serves as a solution for the image morphing task with many applications. In this work, we present DiffMorpher, the first approach enabling smooth and natural image interpolation using diffusion models. Our key idea is to capture the semantics of the two images by fitting two LoRAs to them respectively, and interpolate between both the LoRA parameters and the latent noises to ensure a smooth semantic transition, where correspondence automatically emerges without the need for annotation. In addition, we propose an attention interpolation and injection technique and a new sampling schedule to further enhance the smoothness between consecutive images. Extensive experiments demonstrate that DiffMorpher achieves starkly better image morphing effects than previous methods across a variety of object categories, bridging a critical functional gap that distinguished diffusion models from GANs.
Authors: Kabita Parajuli, Shashidhar Ram Joshi
Video captioning in Nepali, a language written in the Devanagari script, presents a unique challenge due to the lack of existing academic work in this domain. This work develops a novel encoder-decoder paradigm for Nepali video captioning to tackle this difficulty. LSTM and GRU sequence-to-sequence models are used in the model to produce related textual descriptions based on features retrieved from video frames using CNNs. Using Google Translate and manual post-editing, a Nepali video captioning dataset is generated from the Microsoft Research Video Description Corpus (MSVD) dataset created using Google Translate, and manual post-editing work. The efficacy of the model for Devanagari-scripted video captioning is demonstrated by BLEU, METOR, and ROUGE measures, which are used to assess its performance.
Authors: Ashwath Shetty, Marc Habermann, Guoxing Sun, Diogo Luvizon, Vladislav Golyanik, Christian Theobalt
We present the first approach to render highly realistic free-viewpoint videos of a human actor in general apparel, from sparse multi-view recording to display, in real-time at an unprecedented 4K resolution. At inference, our method only requires four camera views of the moving actor and the respective 3D skeletal pose. It handles actors in wide clothing, and reproduces even fine-scale dynamic detail, e.g. clothing wrinkles, face expressions, and hand gestures. At training time, our learning-based approach expects dense multi-view video and a rigged static surface scan of the actor. Our method comprises three main stages. Stage 1 is a skeleton-driven neural approach for high-quality capture of the detailed dynamic mesh geometry. Stage 2 is a novel solution to create a view-dependent texture using four test-time camera views as input. Finally, stage 3 comprises a new image-based refinement network rendering the final 4K image given the output from the previous stages. Our approach establishes a new benchmark for real-time rendering resolution and quality using sparse input camera views, unlocking possibilities for immersive telepresence.
Authors: Zhongyi Han, Guanglin Zhou, Rundong He, Jindong Wang, Xing Xie, Tailin Wu, Yilong Yin, Salman Khan, Lina Yao, Tongliang Liu, Kun Zhang
In machine learning, generalization against distribution shifts -- where deployment conditions diverge from the training scenarios -- is crucial, particularly in fields like climate modeling, biomedicine, and autonomous driving. The emergence of foundation models, distinguished by their extensive pretraining and task versatility, has led to an increased interest in their adaptability to distribution shifts. GPT-4V(ision) acts as the most advanced publicly accessible multimodal foundation model, with extensive applications across various domains, including anomaly detection, video understanding, image generation, and medical diagnosis. However, its robustness against data distributions remains largely underexplored. Addressing this gap, this study rigorously evaluates GPT-4V's adaptability and generalization capabilities in dynamic environments, benchmarking against prominent models like CLIP and LLaVA. We delve into GPT-4V's zero-shot generalization across 13 diverse datasets spanning natural, medical, and molecular domains. We further investigate its adaptability to controlled data perturbations and examine the efficacy of in-context learning as a tool to enhance its adaptation. Our findings delineate GPT-4V's capability boundaries in distribution shifts, shedding light on its strengths and limitations across various scenarios. Importantly, this investigation contributes to our understanding of how AI foundation models generalize to distribution shifts, offering pivotal insights into their adaptability and robustness. Code is publicly available at https://github.com/jameszhou-gl/gpt-4v-distribution-shift.
Authors: Tom Tirer, Raja Giryes, Se Young Chun, Yonina C. Eldar
Deep learning in general focuses on training a neural network from large labeled datasets. Yet, in many cases there is value in training a network just from the input at hand. This may involve training a network from scratch using a single input or adapting an already trained network to a provided input example at inference time. This survey paper aims at covering deep internal-learning techniques that have been proposed in the past few years for these two important directions. While our main focus will be on image processing problems, most of the approaches that we survey are derived for general signals (vectors with recurring patterns that can be distinguished from noise) and are therefore applicable to other modalities. We believe that the topic of internal-learning is very important in many signal and image processing problems where training data is scarce and diversity is large on the one hand, and on the other, there is a lot of structure in the data that can be exploited.
Authors: Alhassan Mabrouk, Rebeca P. Díaz Redondo, Mohamed Abd Elaziz, Mohammed Kayed
Federated learning is a very convenient approach for scenarios where (i) the exchange of data implies privacy concerns and/or (ii) a quick reaction is needed. In smart healthcare systems, both aspects are usually required. In this paper, we work on the first scenario, where preserving privacy is key and, consequently, building a unique and massive medical image data set by fusing different data sets from different medical institutions or research centers (computation nodes) is not an option. We propose an ensemble federated learning (EFL) approach that is based on the following characteristics: First, each computation node works with a different data set (but of the same type). They work locally and apply an ensemble approach combining eight well-known CNN models (densenet169, mobilenetv2, xception, inceptionv3, vgg16, resnet50, densenet121, and resnet152v2) on Chest X-ray images. Second, the best two local models are used to create a local ensemble model that is shared with a central node. Third, the ensemble models are aggregated to obtain a global model, which is shared with the computation nodes to continue with a new iteration. This procedure continues until there are no changes in the best local models. We have performed different experiments to compare our approach with centralized ones (with or without an ensemble approach)\color{black}. The results conclude that our proposal outperforms these ones in Chest X-ray images (achieving an accuracy of 96.63\%) and offers very competitive results compared to other proposals in the literature.
Authors: Love Panta, Prashant Shrestha, Brabeem Sapkota, Amrita Bhattarai, Suresh Manandhar, Anand Kumar Sah
Video moment retrieval is a challenging task requiring fine-grained interactions between video and text modalities. Recent work in image-text pretraining has demonstrated that most existing pretrained models suffer from information asymmetry due to the difference in length between visual and textual sequences. We question whether the same problem also exists in the video-text domain with an auxiliary need to preserve both spatial and temporal information. Thus, we evaluate a recently proposed solution involving the addition of an asymmetric co-attention network for video grounding tasks. Additionally, we incorporate momentum contrastive loss for robust, discriminative representation learning in both modalities. We note that the integration of these supplementary modules yields better performance compared to state-of-the-art models on the TACoS dataset and comparable results on ActivityNet Captions, all while utilizing significantly fewer parameters with respect to baseline.
Authors: Alhassan Mabrouk, Abdelghani Dahou, Mohamed Abd Elaziz, Rebeca P. Díaz Redondo, Mohammed Kayed
The Internet of Medical Things (IoMT) has dramatically benefited medical professionals that patients and physicians can access from all regions. Although the automatic detection and prediction of diseases such as melanoma and leukemia is still being researched and studied in IoMT, existing approaches are not able to achieve a high degree of efficiency. Thus, with a new approach that provides better results, patients would access the adequate treatments earlier and the death rate would be reduced. Therefore, this paper introduces an IoMT proposal for medical images classification that may be used anywhere, i.e. it is an ubiquitous approach. It was design in two stages: first, we employ a Transfer Learning (TL)-based method for feature extraction, which is carried out using MobileNetV3; second, we use the Chaos Game Optimization (CGO) for feature selection, with the aim of excluding unnecessary features and improving the performance, which is key in IoMT. Our methodology was evaluated using ISIC-2016, PH2, and Blood-Cell datasets. The experimental results indicated that the proposed approach obtained an accuracy of 88.39% on ISIC-2016, 97.52% on PH2, and 88.79% on Blood-cell. Moreover, our approach had successful performances for the metrics employed compared to other existing methods.
Authors: Jamil Fayyad, Shadi Alijani, Homayoun Najjaran
Uncertainty quantification is a pivotal field that contributes to the realization of reliable and robust systems. By providing complementary information, it becomes instrumental in fortifying safe decisions, particularly within high-risk applications. Nevertheless, a comprehensive understanding of the advantages and limitations inherent in various methods within the medical imaging field necessitates further research coupled with in-depth analysis. In this paper, we explore Conformal Prediction, an emerging distribution-free uncertainty quantification technique, along with Monte Carlo Dropout and Evidential Deep Learning methods. Our comprehensive experiments provide a comparative performance analysis for skin lesion classification tasks across the three quantification methods. Furthermore, We present insights into the effectiveness of each method in handling Out-of-Distribution samples from domain-shifted datasets. Based on our experimental findings, our conclusion highlights the robustness and consistent performance of conformal prediction across diverse conditions. This positions it as the preferred choice for decision-making in safety-critical applications.
Authors: Aitor Martinez Seras, Javier Del Ser, Pablo Garcia-Bringas
Besides performance, efficiency is a key design driver of technologies supporting vehicular perception. Indeed, a well-balanced trade-off between performance and energy consumption is crucial for the sustainability of autonomous vehicles. In this context, the diversity of real-world contexts in which autonomous vehicles can operate motivates the need for empowering perception models with the capability to detect, characterize and identify newly appearing objects by themselves. In this manuscript we elaborate on this threefold conundrum (performance, efficiency and open-world learning) for object detection modeling tasks over image data collected from vehicular scenarios. Specifically, we show that well-performing and efficient models can be realized by virtue of Spiking Neural Networks (SNNs), reaching competitive levels of detection performance when compared to their non-spiking counterparts at dramatic energy consumption savings (up to 85%) and a slightly improved robustness against image noise. Our experiments herein offered also expose qualitatively the complexity of detecting new objects based on the preliminary results of a simple approach to discriminate potential object proposals in the captured image.
Authors: Yiran Qin, Enshen Zhou, Qichang Liu, Zhenfei Yin, Lu Sheng, Ruimao Zhang, Yu Qiao, Jing Shao
It is a long-lasting goal to design an embodied system that can solve long-horizon open-world tasks in human-like ways. However, existing approaches usually struggle with compound difficulties caused by the logic-aware decomposition and context-aware execution of these tasks. To this end, we introduce MP5, an open-ended multimodal embodied system built upon the challenging Minecraft simulator, which can decompose feasible sub-objectives, design sophisticated situation-aware plans, and perform embodied action control, with frequent communication with a goal-conditioned active perception scheme. Specifically, MP5 is developed on top of recent advances in Multimodal Large Language Models (MLLMs), and the system is modulated into functional modules that can be scheduled and collaborated to ultimately solve pre-defined context- and process-dependent tasks. Extensive experiments prove that MP5 can achieve a 22% success rate on difficult process-dependent tasks and a 91% success rate on tasks that heavily depend on the context. Moreover, MP5 exhibits a remarkable ability to address many open-ended tasks that are entirely novel.
Authors: Zihao Wang, Jing Zhao, Hui Zhang
Face plays an important role in human's visual perception, and reconstructing perceived faces from brain activities is challenging because of its difficulty in extracting high-level features and maintaining consistency of multiple face attributes, such as expression, identity, gender, etc. In this study, we proposed a novel reconstruction framework, which we called Double-Flow GAN, that can enhance the capability of discriminator and handle imbalances in images from certain domains that are too easy for generators. We also designed a pretraining process that uses features extracted from images as conditions for making it possible to pretrain the conditional reconstruction model from fMRI in a larger pure image dataset. Moreover, we developed a simple pretrained model to perform fMRI alignment to alleviate the problem of cross-subject reconstruction due to the variations of brain structure among different subjects. We conducted experiments by using our proposed method and state-of-the-art reconstruction models. Our results demonstrated that our method showed significant reconstruction performance, outperformed the previous reconstruction models, and exhibited a good generation ability.
Authors: Jianxiong Gao, Yuqian Fu, Yun Wang, Xuelin Qian, Jianfeng Feng, Yanwei Fu
In this paper, we introduce Recon3DMind, a groundbreaking task focused on reconstructing 3D visuals from Functional Magnetic Resonance Imaging (fMRI) signals. This represents a major step forward in cognitive neuroscience and computer vision. To support this task, we present the fMRI-Shape dataset, utilizing 360-degree view videos of 3D objects for comprehensive fMRI signal capture. Containing 55 categories of common objects from daily life, this dataset will bolster future research endeavors. We also propose MinD-3D, a novel and effective three-stage framework that decodes and reconstructs the brain's 3D visual information from fMRI signals. This method starts by extracting and aggregating features from fMRI frames using a neuro-fusion encoder, then employs a feature bridge diffusion model to generate corresponding visual features, and ultimately recovers the 3D object through a generative transformer decoder. Our experiments demonstrate that this method effectively extracts features that are valid and highly correlated with visual regions of interest (ROIs) in fMRI signals. Notably, it not only reconstructs 3D objects with high semantic relevance and spatial similarity but also significantly deepens our understanding of the human brain's 3D visual processing capabilities. Project page at: https://jianxgao.github.io/MinD-3D.
Authors: Hao Shao, Yuxuan Hu, Letian Wang, Steven L. Waslander, Yu Liu, Hongsheng Li
Despite significant recent progress in the field of autonomous driving, modern methods still struggle and can incur serious accidents when encountering long-tail unforeseen events and challenging urban scenarios. On the one hand, large language models (LLM) have shown impressive reasoning capabilities that approach "Artificial General Intelligence". On the other hand, previous autonomous driving methods tend to rely on limited-format inputs (e.g. sensor data and navigation waypoints), restricting the vehicle's ability to understand language information and interact with humans. To this end, this paper introduces LMDrive, a novel language-guided, end-to-end, closed-loop autonomous driving framework. LMDrive uniquely processes and integrates multi-modal sensor data with natural language instructions, enabling interaction with humans and navigation software in realistic instructional settings. To facilitate further research in language-based closed-loop autonomous driving, we also publicly release the corresponding dataset which includes approximately 64K instruction-following data clips, and the LangAuto benchmark that tests the system's ability to handle complex instructions and challenging driving scenarios. Extensive closed-loop experiments are conducted to demonstrate LMDrive's effectiveness. To the best of our knowledge, we're the very first work to leverage LLMs for closed-loop end-to-end autonomous driving. Codes can be found at https://github.com/opendilab/LMDrive
Authors: Gia-Bao Le, Van-Tien Nguyen, Trung-Nghia Le, Minh-Triet Tran
Whole-slide image (WSI) analysis plays a crucial role in cancer diagnosis and treatment. In addressing the demands of this critical task, self-supervised learning (SSL) methods have emerged as a valuable resource, leveraging their efficiency in circumventing the need for a large number of annotations, which can be both costly and time-consuming to deploy supervised methods. Nevertheless, patch-wise representation may exhibit instability in performance, primarily due to class imbalances stemming from patch selection within WSIs. In this paper, we introduce Nearby Patch Contrastive Learning (NearbyPatchCL), a novel self-supervised learning method that leverages nearby patches as positive samples and a decoupled contrastive loss for robust representation learning. Our method demonstrates a tangible enhancement in performance for downstream tasks involving patch-level multi-class classification. Additionally, we curate a new dataset derived from WSIs sourced from the Canine Cutaneous Cancer Histology, thus establishing a benchmark for the rigorous evaluation of patch-level multi-class classification methodologies. Intensive experiments show that our method significantly outperforms the supervised baseline and state-of-the-art SSL methods with top-1 classification accuracy of 87.56%. Our method also achieves comparable results while utilizing a mere 1% of labeled data, a stark contrast to the 100% labeled data requirement of other approaches. Source code: https://github.com/nvtien457/NearbyPatchCL
Authors: Han Wu, Syed Yunas, Sareh Rowlands, Wenjie Ruan, Johan Wahlstrom
As research in deep neural networks advances, deep convolutional networks become promising for autonomous driving tasks. In particular, there is an emerging trend of employing end-to-end neural network models for autonomous driving. However, previous research has shown that deep neural network classifiers are vulnerable to adversarial attacks. While for regression tasks, the effect of adversarial attacks is not as well understood. In this research, we devise two white-box targeted attacks against end-to-end autonomous driving models. Our attacks manipulate the behavior of the autonomous driving system by perturbing the input image. In an average of 800 attacks with the same attack strength (epsilon=1), the image-specific and image-agnostic attack deviates the steering angle from the original output by 0.478 and 0.111, respectively, which is much stronger than random noises that only perturbs the steering angle by 0.002 (The steering angle ranges from [-1, 1]). Both attacks can be initiated in real-time on CPUs without employing GPUs. Demo video: https://youtu.be/I0i8uN2oOP0.
Authors: Hengkang Wang, Taihui Li, Zhong Zhuang, Tiancong Chen, Hengyue Liang, Ju Sun
Deep image prior (DIP) and its variants have showed remarkable potential for solving inverse problems in computer vision, without any extra training data. Practical DIP models are often substantially overparameterized. During the fitting process, these models learn mostly the desired visual content first, and then pick up the potential modeling and observational noise, i.e., overfitting. Thus, the practicality of DIP often depends critically on good early stopping (ES) that captures the transition period. In this regard, the majority of DIP works for vision tasks only demonstrates the potential of the models -- reporting the peak performance against the ground truth, but provides no clue about how to operationally obtain near-peak performance without access to the groundtruth. In this paper, we set to break this practicality barrier of DIP, and propose an efficient ES strategy, which consistently detects near-peak performance across several vision tasks and DIP variants. Based on a simple measure of dispersion of consecutive DIP reconstructions, our ES method not only outpaces the existing ones -- which only work in very narrow domains, but also remains effective when combined with a number of methods that try to mitigate the overfitting. The code is available at https://github.com/sun-umn/Early_Stopping_for_DIP.
Authors: Mengwei Ren, Neel Dey, Martin A. Styner, Kelly Botteron, Guido Gerig
Recent self-supervised advances in medical computer vision exploit global and local anatomical self-similarity for pretraining prior to downstream tasks such as segmentation. However, current methods assume i.i.d. image acquisition, which is invalid in clinical study designs where follow-up longitudinal scans track subject-specific temporal changes. Further, existing self-supervised methods for medically-relevant image-to-image architectures exploit only spatial or temporal self-similarity and only do so via a loss applied at a single image-scale, with naive multi-scale spatiotemporal extensions collapsing to degenerate solutions. To these ends, this paper makes two contributions: (1) It presents a local and multi-scale spatiotemporal representation learning method for image-to-image architectures trained on longitudinal images. It exploits the spatiotemporal self-similarity of learned multi-scale intra-subject features for pretraining and develops several feature-wise regularizations that avoid collapsed identity representations; (2) During finetuning, it proposes a surprisingly simple self-supervised segmentation consistency regularization to exploit intra-subject correlation. Benchmarked in the one-shot segmentation setting, the proposed framework outperforms both well-tuned randomly-initialized baselines and current self-supervised techniques designed for both i.i.d. and longitudinal datasets. These improvements are demonstrated across both longitudinal neurodegenerative adult MRI and developing infant brain MRI and yield both higher performance and longitudinal consistency.
Authors: Ramya Muthukrishnan, Angelina Heyler, Keshava Katti, Sarthak Pati, Walter Mankowski, Aprupa Alahari, Michael Sanborn, Emily F. Conant, Christopher Scott, Stacey Winham, Celine Vachon, Pratik Chaudhari, Despina Kontos, Spyridon Bakas
In this study, we automate quantitative mammographic breast density estimation with neural networks and show that this tool is a strong use case for federated learning on multi-institutional datasets. Our dataset included bilateral CC-view and MLO-view mammographic images from two separate institutions. Two U-Nets were separately trained on algorithm-generated labels to perform segmentation of the breast and dense tissue from these images and subsequently calculate breast percent density (PD). The networks were trained with federated learning and compared to three non-federated baselines, one trained on each single-institution dataset and one trained on the aggregated multi-institution dataset. We demonstrate that training on multi-institution datasets is critical to algorithm generalizability. We further show that federated learning on multi-institutional datasets improves model generalization to unseen data at nearly the same level as centralized training on multi-institutional datasets, indicating that federated learning can be applied to our method to improve algorithm generalizability while maintaining patient privacy.
Authors: Han Wu, Syed Yunas, Sareh Rowlands, Wenjie Ruan, Johan Wahlstrom
Intelligent robots rely on object detection models to perceive the environment. Following advances in deep learning security it has been revealed that object detection models are vulnerable to adversarial attacks. However, prior research primarily focuses on attacking static images or offline videos. Therefore, it is still unclear if such attacks could jeopardize real-world robotic applications in dynamic environments. This paper bridges this gap by presenting the first real-time online attack against object detection models. We devise three attacks that fabricate bounding boxes for nonexistent objects at desired locations. The attacks achieve a success rate of about 90% within about 20 iterations. The demo video is available at https://youtu.be/zJZ1aNlXsMU.
Authors: Sen Pei, Shixiong Xu, Xiaojie Jin
Video highlights detection (VHD) is an active research field in computer vision, aiming to locate the most user-appealing clips given raw video inputs. However, most VHD methods are based on the closed world assumption, i.e., a fixed number of highlight categories is defined in advance and all training data are available beforehand. Consequently, existing methods have poor scalability with respect to increasing highlight domains and training data. To address above issues, we propose a novel video highlights detection method named Global Prototype Encoding (GPE) to learn incrementally for adapting to new domains via parameterized prototypes. To facilitate this new research direction, we collect a finely annotated dataset termed LiveFood, including over 5,100 live gourmet videos that consist of four domains: ingredients, cooking, presentation, and eating. To the best of our knowledge, this is the first work to explore video highlights detection in the incremental learning setting, opening up new land to apply VHD for practical scenarios where both the concerned highlight domains and training data increase over time. We demonstrate the effectiveness of GPE through extensive experiments. Notably, GPE surpasses popular domain incremental learning methods on LiveFood, achieving significant mAP improvements on all domains. Concerning the classic datasets, GPE also yields comparable performance as previous arts. The code is available at: https://github.com/ForeverPs/IncrementalVHD_GPE.
Authors: Zhaoyuan Yang, Zhiwei Xu, Jing Zhang, Richard Hartley, Peter Tu
In this work, we formulate a novel framework for adversarial robustness using the manifold hypothesis. This framework provides sufficient conditions for defending against adversarial examples. We develop an adversarial purification method with this framework. Our method combines manifold learning with variational inference to provide adversarial robustness without the need for expensive adversarial training. Experimentally, our approach can provide adversarial robustness even if attackers are aware of the existence of the defense. In addition, our method can also serve as a test-time defense mechanism for variational autoencoders.
Authors: Feifei Wang, Yong Wang, Bing Li, Qidong Huang, Shaoqing Chen
With the deterioration of climate, the phenomenon of rain-induced flooding has become frequent. To mitigate its impact, recent works adopt convolutional neural network or its variants to predict the floods. However, these methods directly force the model to reconstruct the raw pixels of flood images through a global constraint, overlooking the underlying information contained in terrain features and rainfall patterns. To address this, we present a novel framework for precise flood map prediction, which incorporates hierarchical terrain spatial attention to help the model focus on spatially-salient areas of terrain features and constructs multi-scale rainfall embedding to extensively integrate rainfall pattern information into generation. To better adapt the model in various rainfall conditions, we leverage a rainfall regression loss for both the generator and the discriminator as additional supervision. Extensive evaluations on real catchment datasets demonstrate the superior performance of our method, which greatly surpasses the previous arts under different rainfall conditions.
Authors: Tanish Mittal, Preyansh Agrawal, Esha Pahwa, Aarya Makwana
Multi-Scale and U-shaped Networks are widely used in various image restoration problems, including deblurring. Keeping in mind the wide range of applications, we present a comparison of these architectures and their effects on image deblurring. We also introduce a new block called as NFResblock. It consists of a Fast Fourier Transformation layer and a series of modified Non-Linear Activation Free Blocks. Based on these architectures and additions, we introduce NFResnet and NFResnet+, which are modified multi-scale and U-Net architectures, respectively. We also use three different loss functions to train these architectures: Charbonnier Loss, Edge Loss, and Frequency Reconstruction Loss. Extensive experiments on the Deep Video Deblurring dataset, along with ablation studies for each component, have been presented in this paper. The proposed architectures achieve a considerable increase in Peak Signal to Noise (PSNR) ratio and Structural Similarity Index (SSIM) value.
Authors: Jinwoo Bae, Kyumin Hwang, Sunghoon Im
Monocular depth estimation has been widely studied, and significant improvements in performance have been recently reported. However, most previous works are evaluated on a few benchmark datasets, such as KITTI datasets, and none of the works provide an in-depth analysis of the generalization performance of monocular depth estimation. In this paper, we deeply investigate the various backbone networks (e.g.CNN and Transformer models) toward the generalization of monocular depth estimation. First, we evaluate state-of-the-art models on both in-distribution and out-of-distribution datasets, which have never been seen during network training. Then, we investigate the internal properties of the representations from the intermediate layers of CNN-/Transformer-based models using synthetic texture-shifted datasets. Through extensive experiments, we observe that the Transformers exhibit a strong shape-bias rather than CNNs, which have a strong texture-bias. We also discover that texture-biased models exhibit worse generalization performance for monocular depth estimation than shape-biased models. We demonstrate that similar aspects are observed in real-world driving datasets captured under diverse environments. Lastly, we conduct a dense ablation study with various backbone networks which are utilized in modern strategies. The experiments demonstrate that the intrinsic locality of the CNNs and the self-attention of the Transformers induce texture-bias and shape-bias, respectively.
Authors: Mohammadreza Shetab-Bushehri, Miguel Aranda, Youcef Mezouar, Adrien Bartoli, Erol Ozgur
Tracking the 3D shape of a deforming object using only monocular 2D vision is a challenging problem. This is because one should (i) infer the 3D shape from a 2D image, which is a severely underconstrained problem, and (ii) implement the whole solution pipeline in real-time. The pipeline typically requires feature detection and matching, mismatch filtering, 3D shape inference and feature tracking algorithms. We propose ROBUSfT, a conventional pipeline based on a template containing the object's rest shape, texturemap and deformation law. ROBUSfT is ready-to-use, wide-baseline, capable of handling large deformations, fast up to 30 fps, free of training, and robust against partial occlusions and discontinuity in video frames. It outperforms the state-of-the-art methods in challenging datasets. ROBUSfT is implemented as a publicly available C++ library and we provide a tutorial on how to use it in https://github.com/mrshetab/ROBUSfT
Authors: Emiel Hoogeboom, Jonathan Heek, Tim Salimans
Currently, applying diffusion models in pixel space of high resolution images is difficult. Instead, existing approaches focus on diffusion in lower dimensional spaces (latent diffusion), or have multiple super-resolution levels of generation referred to as cascades. The downside is that these approaches add additional complexity to the diffusion framework.
This paper aims to improve denoising diffusion for high resolution images while keeping the model as simple as possible. The paper is centered around the research question: How can one train a standard denoising diffusion models on high resolution images, and still obtain performance comparable to these alternate approaches?
The four main findings are: 1) the noise schedule should be adjusted for high resolution images, 2) It is sufficient to scale only a particular part of the architecture, 3) dropout should be added at specific locations in the architecture, and 4) downsampling is an effective strategy to avoid high resolution feature maps. Combining these simple yet effective techniques, we achieve state-of-the-art on image generation among diffusion models without sampling modifiers on ImageNet.
Authors: Wenlong Huang, Fei Xia, Dhruv Shah, Danny Driess, Andy Zeng, Yao Lu, Pete Florence, Igor Mordatch, Sergey Levine, Karol Hausman, Brian Ichter
Recent progress in large language models (LLMs) has demonstrated the ability to learn and leverage Internet-scale knowledge through pre-training with autoregressive models. Unfortunately, applying such models to settings with embodied agents, such as robots, is challenging due to their lack of experience with the physical world, inability to parse non-language observations, and ignorance of rewards or safety constraints that robots may require. On the other hand, language-conditioned robotic policies that learn from interaction data can provide the necessary grounding that allows the agent to be correctly situated in the real world, but such policies are limited by the lack of high-level semantic understanding due to the limited breadth of the interaction data available for training them. Thus, if we want to make use of the semantic knowledge in a language model while still situating it in an embodied setting, we must construct an action sequence that is both likely according to the language model and also realizable according to grounded models of the environment. We frame this as a problem similar to probabilistic filtering: decode a sequence that both has high probability under the language model and high probability under a set of grounded model objectives. We demonstrate how such grounded models can be obtained across three simulation and real-world domains, and that the proposed decoding strategy is able to solve complex, long-horizon embodiment tasks in a robotic setting by leveraging the knowledge of both models. The project's website can be found at grounded-decoding.github.io.
Authors: Xuyang Zhao, Tianqi Du, Yisen Wang, Jun Yao, Weiran Huang
Self-Supervised Learning (SSL) is a paradigm that leverages unlabeled data for model training. Empirical studies show that SSL can achieve promising performance in distribution shift scenarios, where the downstream and training distributions differ. However, the theoretical understanding of its transferability remains limited. In this paper, we develop a theoretical framework to analyze the transferability of self-supervised contrastive learning, by investigating the impact of data augmentation on it. Our results reveal that the downstream performance of contrastive learning depends largely on the choice of data augmentation. Moreover, we show that contrastive learning fails to learn domain-invariant features, which limits its transferability. Based on these theoretical insights, we propose a novel method called Augmentation-robust Contrastive Learning (ArCL), which guarantees to learn domain-invariant features and can be easily integrated with existing contrastive learning algorithms. We conduct experiments on several datasets and show that ArCL significantly improves the transferability of contrastive learning.
Authors: Tianqi Wang, Sukmin Kim, Wenxuan Ji, Enze Xie, Chongjian Ge, Junsong Chen, Zhenguo Li, Ping Luo
Safety is the primary priority of autonomous driving. Nevertheless, no published dataset currently supports the direct and explainable safety evaluation for autonomous driving. In this work, we propose DeepAccident, a large-scale dataset generated via a realistic simulator containing diverse accident scenarios that frequently occur in real-world driving. The proposed DeepAccident dataset includes 57K annotated frames and 285K annotated samples, approximately 7 times more than the large-scale nuScenes dataset with 40k annotated samples. In addition, we propose a new task, end-to-end motion and accident prediction, which can be used to directly evaluate the accident prediction ability for different autonomous driving algorithms. Furthermore, for each scenario, we set four vehicles along with one infrastructure to record data, thus providing diverse viewpoints for accident scenarios and enabling V2X (vehicle-to-everything) research on perception and prediction tasks. Finally, we present a baseline V2X model named V2XFormer that demonstrates superior performance for motion and accident prediction and 3D object detection compared to the single-vehicle model.
Authors: Hongbo Tian, Yulong Li, Linzhi Huang, Xu Ling, Yue Yang, Jiani Hu
Structured reconstruction is a non-trivial dense prediction problem, which extracts structural information (\eg, building corners and edges) from a raster image, then reconstructs it to a 2D planar graph accordingly. Compared with common segmentation or detection problems, it significantly relays on the capability that leveraging holistic geometric information for structural reasoning. Current transformer-based approaches tackle this challenging problem in a two-stage manner, which detect corners in the first model and classify the proposed edges (corner-pairs) in the second model. However, they separate two-stage into different models and only share the backbone encoder. Unlike the existing modeling strategies, we present an enhanced corner representation method: 1) It fuses knowledge between the corner detection and edge prediction by sharing feature in different granularity; 2) Corner candidates are proposed in four heatmap channels w.r.t its direction. Both qualitative and quantitative evaluations demonstrate that our proposed method can better reconstruct fine-grained structures, such as adjacent corners and tiny edges. Consequently, it outperforms the state-of-the-art model by +1.9\%@F-1 on Corner and +3.0\%@F-1 on Edge.
Authors: Jannis Priesnitz, Axel Weißenfeld, Laurenz Ruzicka, Christian Rathgeb, Bernhard Strobl, Ralph Lessmann, Christoph Busch
We propose MCLFIQ: Mobile Contactless Fingerprint Image Quality, the first quality assessment algorithm for mobile contactless fingerprint samples. To this end, we re-trained the NIST Fingerprint Image Quality (NFIQ) 2 method, which was originally designed for contact-based fingerprints, with a synthetic contactless fingerprint database. We evaluate the predictive performance of the resulting MCLFIQ model in terms of Error-vs.-Discard Characteristic (EDC) curves on three real-world contactless fingerprint databases using three recognition algorithms. In experiments, the MCLFIQ method is compared against the original NFIQ 2.2 method, a sharpness-based quality assessment algorithm developed for contactless fingerprint images \rev{and the general purpose image quality assessment method BRISQUE. Furthermore, benchmarks on four contact-based fingerprint datasets are also conducted.}
Obtained results show that the fine-tuning of NFIQ 2 on synthetic contactless fingerprints is a viable alternative to training on real databases. Moreover, the evaluation shows that our MCLFIQ method works more accurate and robust compared to all baseline methods on contactless fingerprints. We suggest considering the proposed MCLFIQ method as a \rev{starting point for the development of} a new standard algorithm for contactless fingerprint quality assessment.
Authors: Yuhao Huang, Xin Yang, Lian Liu, Han Zhou, Ao Chang, Xinrui Zhou, Rusi Chen, Junxuan Yu, Jiongquan Chen, Chaoyu Chen, Sijing Liu, Haozhe Chi, Xindi Hu, Kejuan Yue, Lei Li, Vicente Grau, Deng-Ping Fan, Fajin Dong, Dong Ni
The Segment Anything Model (SAM) is the first foundation model for general image segmentation. It has achieved impressive results on various natural image segmentation tasks. However, medical image segmentation (MIS) is more challenging because of the complex modalities, fine anatomical structures, uncertain and complex object boundaries, and wide-range object scales. To fully validate SAM's performance on medical data, we collected and sorted 53 open-source datasets and built a large medical segmentation dataset with 18 modalities, 84 objects, 125 object-modality paired targets, 1050K 2D images, and 6033K masks. We comprehensively analyzed different models and strategies on the so-called COSMOS 1050K dataset. Our findings mainly include the following: 1) SAM showed remarkable performance in some specific objects but was unstable, imperfect, or even totally failed in other situations. 2) SAM with the large ViT-H showed better overall performance than that with the small ViT-B. 3) SAM performed better with manual hints, especially box, than the Everything mode. 4) SAM could help human annotation with high labeling quality and less time. 5) SAM was sensitive to the randomness in the center point and tight box prompts, and may suffer from a serious performance drop. 6) SAM performed better than interactive methods with one or a few points, but will be outpaced as the number of points increases. 7) SAM's performance correlated to different factors, including boundary complexity, intensity differences, etc. 8) Finetuning the SAM on specific medical tasks could improve its average DICE performance by 4.39% and 6.68% for ViT-B and ViT-H, respectively. We hope that this comprehensive report can help researchers explore the potential of SAM applications in MIS, and guide how to appropriately use and develop SAM.
Authors: Subin Lin, Vasantha Ramani, Miguel Martin, Pandarasamy Arjunan, Adrian Chong, Filip Biljecki, Marcel Ignatius, Kameshwar Poolla, Clayton Miller
The paper describes a dataset that was collected by infrared thermography, which is a non-contact, non-intrusive technique to collect data and analyze the built environment in various aspects. While most studies focus on the city and building scales, the rooftop observatory provides high temporal and spatial resolution observations with dynamic interactions on the district scale. The rooftop infrared thermography observatory with a multi-modal platform that is capable of assessing a wide range of dynamic processes in urban systems was deployed in Singapore. It was placed on the top of two buildings that overlook the outdoor context of the campus of the National University of Singapore. The platform collects remote sensing data from tropical areas on a temporal scale, allowing users to determine the temperature trend of individual features such as buildings, roads, and vegetation. The dataset includes 1,365,921 thermal images collected on average at approximately 10 seconds intervals from two locations during ten months.
Authors: Omer Belhasin, Yaniv Romano, Daniel Freedman, Ehud Rivlin, Michael Elad
Uncertainty quantification for inverse problems in imaging has drawn much attention lately. Existing approaches towards this task define uncertainty regions based on probable values per pixel, while ignoring spatial correlations within the image, resulting in an exaggerated volume of uncertainty. In this paper, we propose PUQ (Principal Uncertainty Quantification) -- a novel definition and corresponding analysis of uncertainty regions that takes into account spatial relationships within the image, thus providing reduced volume regions. Using recent advancements in generative models, we derive uncertainty intervals around principal components of the empirical posterior distribution, forming an ambiguity region that guarantees the inclusion of true unseen values with a user-defined confidence probability. To improve computational efficiency and interpretability, we also guarantee the recovery of true unseen values using only a few principal directions, resulting in more informative uncertainty regions. Our approach is verified through experiments on image colorization, super-resolution, and inpainting; its effectiveness is shown through comparison to baseline methods, demonstrating significantly tighter uncertainty regions.
Authors: Johan Edstedt, Qiyu Sun, Georg Bökman, Mårten Wadenbäck, Michael Felsberg
Feature matching is an important computer vision task that involves estimating correspondences between two images of a 3D scene, and dense methods estimate all such correspondences. The aim is to learn a robust model, i.e., a model able to match under challenging real-world changes. In this work, we propose such a model, leveraging frozen pretrained features from the foundation model DINOv2. Although these features are significantly more robust than local features trained from scratch, they are inherently coarse. We therefore combine them with specialized ConvNet fine features, creating a precisely localizable feature pyramid. To further improve robustness, we propose a tailored transformer match decoder that predicts anchor probabilities, which enables it to express multimodality. Finally, we propose an improved loss formulation through regression-by-classification with subsequent robust regression. We conduct a comprehensive set of experiments that show that our method, RoMa, achieves significant gains, setting a new state-of-the-art. In particular, we achieve a 36% improvement on the extremely challenging WxBS benchmark. Code is provided at https://github.com/Parskatt/RoMa
Authors: Shilin Yan, Renrui Zhang, Ziyu Guo, Wenchao Chen, Wei Zhang, Hongyang Li, Yu Qiao, Hao Dong, Zhongjiang He, Peng Gao
Recently, video object segmentation (VOS) referred by multi-modal signals, e.g., language and audio, has evoked increasing attention in both industry and academia. It is challenging for exploring the semantic alignment within modalities and the visual correspondence across frames. However, existing methods adopt separate network architectures for different modalities, and neglect the inter-frame temporal interaction with references. In this paper, we propose MUTR, a Multi-modal Unified Temporal transformer for Referring video object segmentation. With a unified framework for the first time, MUTR adopts a DETR-style transformer and is capable of segmenting video objects designated by either text or audio reference. Specifically, we introduce two strategies to fully explore the temporal relations between videos and multi-modal signals. Firstly, for low-level temporal aggregation before the transformer, we enable the multi-modal references to capture multi-scale visual cues from consecutive video frames. This effectively endows the text or audio signals with temporal knowledge and boosts the semantic alignment between modalities. Secondly, for high-level temporal interaction after the transformer, we conduct inter-frame feature communication for different object embeddings, contributing to better object-wise correspondence for tracking along the video. On Ref-YouTube-VOS and AVSBench datasets with respective text and audio references, MUTR achieves +4.2% and +8.7% J&F improvements to state-of-the-art methods, demonstrating our significance for unified multi-modal VOS. Code is released at https://github.com/OpenGVLab/MUTR.
Authors: Marawan Elbatel, Robert Martí, Xiaomeng Li
Representational transfer from publicly available models is a promising technique for improving medical image classification, especially in long-tailed datasets with rare diseases. However, existing methods often overlook the frequency-dependent behavior of these models, thereby limiting their effectiveness in transferring representations and generalizations to rare diseases. In this paper, we propose FoPro-KD, a novel framework that leverages the power of frequency patterns learned from frozen pre-trained models to enhance their transferability and compression, presenting a few unique insights: 1) We demonstrate that leveraging representations from publicly available pre-trained models can substantially improve performance, specifically for rare classes, even when utilizing representations from a smaller pre-trained model. 2) We observe that pre-trained models exhibit frequency preferences, which we explore using our proposed Fourier Prompt Generator (FPG), allowing us to manipulate specific frequencies in the input image, enhancing the discriminative representational transfer. 3) By amplifying or diminishing these frequencies in the input image, we enable Effective Knowledge Distillation (EKD). EKD facilitates the transfer of knowledge from pre-trained models to smaller models. Through extensive experiments in long-tailed gastrointestinal image recognition and skin lesion classification, where rare diseases are prevalent, our FoPro-KD framework outperforms existing methods, enabling more accessible medical models for rare disease classification. Code is available at https://github.com/xmed-lab/FoPro-KD.
Authors: Sergey N. Pozdnyakov, Michele Ceriotti
Point clouds are versatile representations of 3D objects and have found widespread application in science and engineering. Many successful deep-learning models have been proposed that use them as input. The domain of chemical and materials modeling is especially challenging because exact compliance with physical constraints is highly desirable for a model to be usable in practice. These constraints include smoothness and invariance with respect to translations, rotations, and permutations of identical atoms. If these requirements are not rigorously fulfilled, atomistic simulations might lead to absurd outcomes even if the model has excellent accuracy. Consequently, dedicated architectures, which achieve invariance by restricting their design space, have been developed. General-purpose point-cloud models are more varied but often disregard rotational symmetry. We propose a general symmetrization method that adds rotational equivariance to any given model while preserving all the other requirements. Our approach simplifies the development of better atomic-scale ML schemes by relaxing the constraints on the design space and making it possible to incorporate ideas that proved effective in other domains. We demonstrate this idea by introducing the Point Edge Transformer (PET) architecture, which is not intrinsically equivariant but achieves state-of-the-art performance on several benchmark datasets of molecules and solids. A-posteriori application of our general protocol makes PET exactly equivariant, with minimal changes to its accuracy.
Authors: Lu Sang, Abhishek Saroha, Maolin Gao, Daniel Cremers
Neural implicits have become popular for representing surfaces because they offer an adaptive resolution and support arbitrary topologies. While previous works rely on ground truth point clouds, they often ignore the effect of input quality and sampling methods during reconstructing process. In this paper, we introduce a sampling method with an uncertainty-augmented surface implicit representation that employs a sampling technique that considers the geometric characteristics of inputs. To this end, we introduce a strategy that efficiently computes differentiable geometric features, namely, mean curvatures, to augment the sampling phase during the training period. The uncertainty augmentation offers insights into the occupancy and reliability of the output signed distance value, thereby expanding representation capabilities into open surfaces. Finally, we demonstrate that our method leads to state-of-the-art reconstructions on both synthetic and real-world data.
Authors: Lingfeng Yang, Yueze Wang, Xiang Li, Xinlong Wang, Jian Yang
Vision-Language Models (VLMs), such as CLIP, have demonstrated impressive zero-shot transfer capabilities in image-level visual perception. However, these models have shown limited performance in instance-level tasks that demand precise localization and recognition. Previous works have suggested that incorporating visual prompts, such as colorful boxes or circles, can improve the ability of models to recognize objects of interest. Nonetheless, compared to language prompting, visual prompting designs are rarely explored. Existing approaches, which employ coarse visual cues such as colorful boxes or circles, often result in sub-optimal performance due to the inclusion of irrelevant and noisy pixels. In this paper, we carefully study the visual prompting designs by exploring more fine-grained markings, such as segmentation masks and their variations. In addition, we introduce a new zero-shot framework that leverages pixel-level annotations acquired from a generalist segmentation model for fine-grained visual prompting. Consequently, our investigation reveals that a straightforward application of blur outside the target mask, referred to as the Blur Reverse Mask, exhibits exceptional effectiveness. This proposed prompting strategy leverages the precise mask annotations to reduce focus on weakly related regions while retaining spatial coherence between the target and the surrounding background. Our Fine-Grained Visual Prompting (FGVP) demonstrates superior performance in zero-shot comprehension of referring expressions on the RefCOCO, RefCOCO+, and RefCOCOg benchmarks. It outperforms prior methods by an average margin of 3.0% to 4.6%, with a maximum improvement of 12.5% on the RefCOCO+ testA subset. Code is available at https://github.com/ylingfeng/FGVP.
Authors: Caixin Kang, Yinpeng Dong, Zhengyi Wang, Shouwei Ruan, Yubo Chen, Hang Su, Xingxing Wei
Adversarial attacks, particularly patch attacks, pose significant threats to the robustness and reliability of deep learning models. Developing reliable defenses against patch attacks is crucial for real-world applications, yet current research in this area is unsatisfactory. In this paper, we propose DIFFender, a novel defense method that leverages a text-guided diffusion model to defend against adversarial patches. DIFFender includes two main stages: patch localization and patch restoration. In the localization stage, we find and exploit an intriguing property of the diffusion model to precisely identify the locations of adversarial patches. In the restoration stage, we employ the diffusion model to reconstruct the adversarial regions in the images while preserving the integrity of the visual content. Thanks to the former finding, these two stages can be simultaneously guided by a unified diffusion model. Thus, we can utilize the close interaction between them to improve the whole defense performance. Moreover, we propose a few-shot prompt-tuning algorithm to fine-tune the diffusion model, enabling the pre-trained diffusion model to adapt to the defense task easily. We conduct extensive experiments on image classification, face recognition, and further in the physical world, demonstrating that our proposed method exhibits superior robustness under strong adaptive attacks and generalizes well across various scenarios, diverse classifiers, and multiple patch attack methods.
Authors: Wentao Liu, Tong Tian, Lemeng Wang, Weijin Xu, Haoyuan Li, Wenyi Zhao, Siyu Tian, Xipeng Pan, Huihua Yang, Feng Gao, Yiming Deng, Ruisheng Su
Digital subtraction angiography (DSA) is universally acknowledged as the gold standard for examining lesion angioarchitecture, elucidating arterial blood supply dynamics, and guiding endovascular interventions. The automatic segmentation of intracranial arteries (IA) in DSA, which is pivotal for quantifying vascular morphology, plays an essential role in computer-assisted stroke research and clinical practices. Nevertheless, research in this specific domain remains constrained, primarily owing to the unavailability of publicly datasets for IA segmentation within the research community. Currently, the predominant focus of methodologies lies in the segmentation of single-frame DSA using in-house datasets. These methods, limited by the partial inclusion of contrast in single-frame DSA, encounters challenges in rendering a precise representation of vascular structures. In this paper, we introduces DIAS, a dataset specifically developed for IA segmentation in DSA sequences. A comprehensive benchmark has been established for evaluating DIAS, covering fully, weakly, and semi-supervised segmentation methods. Specifically, we propose a vessel sequence segmentation network that captures the spatiotemporal representation of intravascular contrast for segmenting vessels in DSA sequences. For weakly-supervised learning, we propose a novel scribble learning-based image segmentation framework, incorporating both scribble supervision and consistency regularization. Furthermore, we introduce a random patch-based self-training framework that harnesses unlabeled DSA sequences to improve segmentation performance. Our extensive experiments on the DIAS dataset demonstrate the effectiveness of these methods as potential baselines for future research and clinical applications.
Authors: Siobhan Mackenzie Hall, Fernanda Gonçalves Abrantes, Hanwen Zhu, Grace Sodunke, Aleksandar Shtedritski, Hannah Rose Kirk
We introduce VisoGender, a novel dataset for benchmarking gender bias in vision-language models. We focus on occupation-related biases within a hegemonic system of binary gender, inspired by Winograd and Winogender schemas, where each image is associated with a caption containing a pronoun relationship of subjects and objects in the scene. VisoGender is balanced by gender representation in professional roles, supporting bias evaluation in two ways: i) resolution bias, where we evaluate the difference between pronoun resolution accuracies for image subjects with gender presentations perceived as masculine versus feminine by human annotators and ii) retrieval bias, where we compare ratios of professionals perceived to have masculine and feminine gender presentations retrieved for a gender-neutral search query. We benchmark several state-of-the-art vision-language models and find that they demonstrate bias in resolving binary gender in complex scenes. While the direction and magnitude of gender bias depends on the task and the model being evaluated, captioning models are generally less biased than Vision-Language Encoders. Dataset and code are available at https://github.com/oxai/visogender
Authors: Yujun Shi, Chuhui Xue, Jun Hao Liew, Jiachun Pan, Hanshu Yan, Wenqing Zhang, Vincent Y. F. Tan, Song Bai
Accurate and controllable image editing is a challenging task that has attracted significant attention recently. Notably, DragGAN is an interactive point-based image editing framework that achieves impressive editing results with pixel-level precision. However, due to its reliance on generative adversarial networks (GANs), its generality is limited by the capacity of pretrained GAN models. In this work, we extend this editing framework to diffusion models and propose a novel approach DragDiffusion. By harnessing large-scale pretrained diffusion models, we greatly enhance the applicability of interactive point-based editing on both real and diffusion-generated images. Our approach involves optimizing the diffusion latents to achieve precise spatial control. The supervision signal of this optimization process is from the diffusion model's UNet features, which are known to contain rich semantic and geometric information. Moreover, we introduce two additional techniques, namely LoRA fine-tuning and latent-MasaCtrl, to further preserve the identity of the original image. Lastly, we present a challenging benchmark dataset called DragBench -- the first benchmark to evaluate the performance of interactive point-based image editing methods. Experiments across a wide range of challenging cases (e.g., images with multiple objects, diverse object categories, various styles, etc.) demonstrate the versatility and generality of DragDiffusion. Code: https://github.com/Yujun-Shi/DragDiffusion.
Authors: Koustava Goswami, Srikrishna Karanam, Prateksha Udhayanan, K J Joseph, Balaji Vasan Srinivasan
Recent advances in multimodal learning has resulted in powerful vision-language models, whose representations are generalizable across a variety of downstream tasks. Recently, their generalization ability has been further extended by incorporating trainable prompts, borrowed from the natural language processing literature. While such prompt learning techniques have shown impressive results, we identify that these prompts are trained based on global image features which limits itself in two aspects: First, by using global features, these prompts could be focusing less on the discriminative foreground image, resulting in poor generalization to various out-of-distribution test cases. Second, existing work weights all prompts equally whereas intuitively, prompts should be reweighed according to the semantics of the image. We address these as part of our proposed Contextual Prompt Learning (CoPL) framework, capable of aligning the prompts to the localized features of the image. Our key innovations over earlier works include using local image features as part of the prompt learning process, and more crucially, learning to weight these prompts based on local features that are appropriate for the task at hand. This gives us dynamic prompts that are both aligned to local image features as well as aware of local contextual relationships. Our extensive set of experiments on a variety of standard and few-shot datasets show that our method produces substantially improved performance when compared to the current state of the art methods. We also demonstrate both few-shot and out-of-distribution performance to establish the utility of learning dynamic prompts that are aligned to local image features.
Authors: Lukas Haas, Michal Skreta, Silas Alberti, Chelsea Finn
Planet-scale image geolocalization remains a challenging problem due to the diversity of images originating from anywhere in the world. Although approaches based on vision transformers have made significant progress in geolocalization accuracy, success in prior literature is constrained to narrow distributions of images of landmarks, and performance has not generalized to unseen places. We present a new geolocalization system that combines semantic geocell creation, multi-task contrastive pretraining, and a novel loss function. Additionally, our work is the first to perform retrieval over location clusters for guess refinements. We train two models for evaluations on street-level data and general-purpose image geolocalization; the first model, PIGEON, is trained on data from the game of Geoguessr and is capable of placing over 40% of its guesses within 25 kilometers of the target location globally. We also develop a bot and deploy PIGEON in a blind experiment against humans, ranking in the top 0.01% of players. We further challenge one of the world's foremost professional Geoguessr players to a series of six matches with millions of viewers, winning all six games. Our second model, PIGEOTTO, differs in that it is trained on a dataset of images from Flickr and Wikipedia, achieving state-of-the-art results on a wide range of image geolocalization benchmarks, outperforming the previous SOTA by up to 7.7 percentage points on the city accuracy level and up to 38.8 percentage points on the country level. Our findings suggest that PIGEOTTO is the first image geolocalization model that effectively generalizes to unseen places and that our approach can pave the way for highly accurate, planet-scale image geolocalization systems. Our code is available on GitHub.
Authors: Xinyu Gao, Ziyi Yang, Yunlu Zhao, Yuxiang Sun, Xiaogang Jin, Changqing Zou
A variety of Neural Radiance Fields (NeRF) methods have recently achieved remarkable success in high render speed. However, current accelerating methods are specialized and incompatible with various implicit methods, preventing real-time composition over various types of NeRF works. Because NeRF relies on sampling along rays, it is possible to provide general guidance for acceleration. To that end, we propose a general implicit pipeline for composing NeRF objects quickly. Our method enables the casting of dynamic shadows within or between objects using analytical light sources while allowing multiple NeRF objects to be seamlessly placed and rendered together with any arbitrary rigid transformations. Mainly, our work introduces a new surface representation known as Neural Depth Fields (NeDF) that quickly determines the spatial relationship between objects by allowing direct intersection computation between rays and implicit surfaces. It leverages an intersection neural network to query NeRF for acceleration instead of depending on an explicit spatial structure.Our proposed method is the first to enable both the progressive and interactive composition of NeRF objects. Additionally, it also serves as a previewing plugin for a range of existing NeRF works.
Authors: Dibyadip Chatterjee, Fadime Sener, Shugao Ma, Angela Yao
Human actions in egocentric videos are often hand-object interactions composed from a verb (performed by the hand) applied to an object. Despite their extensive scaling up, egocentric datasets still face two limitations - sparsity of action compositions and a closed set of interacting objects. This paper proposes a novel open vocabulary action recognition task. Given a set of verbs and objects observed during training, the goal is to generalize the verbs to an open vocabulary of actions with seen and novel objects. To this end, we decouple the verb and object predictions via an object-agnostic verb encoder and a prompt-based object encoder. The prompting leverages CLIP representations to predict an open vocabulary of interacting objects. We create open vocabulary benchmarks on the EPIC-KITCHENS-100 and Assembly101 datasets; whereas closed-action methods fail to generalize, our proposed method is effective. In addition, our object encoder significantly outperforms existing open-vocabulary visual recognition methods in recognizing novel interacting objects.
Authors: Jianning Li, Zongwei Zhou, Jiancheng Yang, Antonio Pepe, Christina Gsaxner, Gijs Luijten, Chongyu Qu, Tiezheng Zhang, Xiaoxi Chen, Wenxuan Li, Marek Wodzinski, Paul Friedrich, Kangxian Xie, Yuan Jin, Narmada Ambigapathy, Enrico Nasca, Naida Solak, Gian Marco Melito, Viet Duc Vu, Afaque R. Memon, Christopher Schlachta, Sandrine De Ribaupierre, Rajnikant Patel, Roy Eagleson, Xiaojun Chen, Heinrich Mächler, Jan Stefan Kirschke, Ezequiel de la Rosa, Patrick Ferdinand Christ, Hongwei Bran Li, David G. Ellis, Michele R. Aizenberg, Sergios Gatidis, Thomas Küstner, Nadya Shusharina, Nicholas Heller, Vincent Andrearczyk, Adrien Depeursinge, Mathieu Hatt, Anjany Sekuboyina, Maximilian Löffler, Hans Liebl, Reuben Dorent, Tom Vercauteren, Jonathan Shapey, Aaron Kujawa, Stefan Cornelissen, et al. (110 additional authors not shown)
Prior to the deep learning era, shape was commonly used to describe the objects. Nowadays, state-of-the-art (SOTA) algorithms in medical imaging are predominantly diverging from computer vision, where voxel grids, meshes, point clouds, and implicit surface models are used. This is seen from numerous shape-related publications in premier vision conferences as well as the growing popularity of ShapeNet (about 51,300 models) and Princeton ModelNet (127,915 models). For the medical domain, we present a large collection of anatomical shapes (e.g., bones, organs, vessels) and 3D models of surgical instrument, called MedShapeNet, created to facilitate the translation of data-driven vision algorithms to medical applications and to adapt SOTA vision algorithms to medical problems. As a unique feature, we directly model the majority of shapes on the imaging data of real patients. As of today, MedShapeNet includes 23 dataset with more than 100,000 shapes that are paired with annotations (ground truth). Our data is freely accessible via a web interface and a Python application programming interface (API) and can be used for discriminative, reconstructive, and variational benchmarks as well as various applications in virtual, augmented, or mixed reality, and 3D printing. Exemplary, we present use cases in the fields of classification of brain tumors, facial and skull reconstructions, multi-class anatomy completion, education, and 3D printing. In future, we will extend the data and improve the interfaces. The project pages are: https://medshapenet.ikim.nrw/ and https://github.com/Jianningli/medshapenet-feedback
Authors: Chen Li, Jiahao Lin, Gim Hee Lee
In this paper, we tackle the challenging task of learning a generalizable human NeRF model from a monocular video. Although existing generalizable human NeRFs have achieved impressive results, they require muti-view images or videos which might not be always available. On the other hand, some works on free-viewpoint rendering of human from monocular videos cannot be generalized to unseen identities. In view of these limitations, we propose GHuNeRF to learn a generalizable human NeRF model from a monocular video of the human performer. We first introduce a visibility-aware aggregation scheme to compute vertex-wise features, which is used to construct a 3D feature volume. The feature volume can only represent the overall geometry of the human performer with insufficient accuracy due to the limited resolution. To solve this, we further enhance the volume feature with temporally aligned point-wise features using an attention mechanism. Finally, the enhanced feature is used for predicting density and color for each sampled point. A surface-guided sampling strategy is also adopted to improve the efficiency for both training and inference. We validate our approach on the widely-used ZJU-MoCap dataset, where we achieve comparable performance with existing multi-view video based approaches. We also test on the monocular People-Snapshot dataset and achieve better performance than existing works when only monocular video is used. Our code is available at the project website.
Authors: Jiaben Chen, Huaizu Jiang
Human-centric video frame interpolation has great potential for improving people's entertainment experiences and finding commercial applications in the sports analysis industry, e.g., synthesizing slow-motion videos. Although there are multiple benchmark datasets available in the community, none of them is dedicated for human-centric scenarios. To bridge this gap, we introduce SportsSloMo, a benchmark consisting of more than 130K video clips and 1M video frames of high-resolution ($\geq$720p) slow-motion sports videos crawled from YouTube. We re-train several state-of-the-art methods on our benchmark, and the results show a decrease in their accuracy compared to other datasets. It highlights the difficulty of our benchmark and suggests that it poses significant challenges even for the best-performing methods, as human bodies are highly deformable and occlusions are frequent in sports videos. To improve the accuracy, we introduce two loss terms considering the human-aware priors, where we add auxiliary supervision to panoptic segmentation and human keypoints detection, respectively. The loss terms are model agnostic and can be easily plugged into any video frame interpolation approaches. Experimental results validate the effectiveness of our proposed loss terms, leading to consistent performance improvement over 5 existing models, which establish strong baseline models on our benchmark. The dataset and code can be found at: https://neu-vi.github.io/SportsSlomo/.
Authors: Jongwon Lee, Su Yeon Choi, Timothy Bretl
This paper quantifies the performance of visual SLAM that leverages multi-scale fiducial markers (i.e., artificial landmarks that can be detected at a wide range of distances) to show its potential for reliable takeoff and landing navigation in rotorcraft. Prior work has shown that square markers with a black-and-white pattern of grid cells can be used to improve the performance of visual SLAM with color cameras. We extend this prior work to allow nested marker layouts. We evaluate performance during semi-autonomous takeoff and landing operations in a variety of environmental conditions by a DJI Matrice 300 RTK rotorcraft with two FLIR Blackfly color cameras, using RTK GNSS to obtain ground truth pose estimates. Performance measures include absolute trajectory error and the fraction of the number of estimated poses to the total frame. We release all of our results -- our dataset and the code of the implementation of the visual SLAM with fiducial markers -- to the public as open-source.
Authors: Shaheer Mohamed, Maryam Haghighat, Tharindu Fernando, Sridha Sridharan, Clinton Fookes, Peyman Moghadam
Hyperspectral images (HSIs) contain rich spectral and spatial information. Motivated by the success of transformers in the field of natural language processing and computer vision where they have shown the ability to learn long range dependencies within input data, recent research has focused on using transformers for HSIs. However, current state-of-the-art hyperspectral transformers only tokenize the input HSI sample along the spectral dimension, resulting in the under-utilization of spatial information. Moreover, transformers are known to be data-hungry and their performance relies heavily on large-scale pre-training, which is challenging due to limited annotated hyperspectral data. Therefore, the full potential of HSI transformers has not been fully realized. To overcome these limitations, we propose a novel factorized spectral-spatial transformer that incorporates factorized self-supervised pre-training procedures, leading to significant improvements in performance. The factorization of the inputs allows the spectral and spatial transformers to better capture the interactions within the hyperspectral data cubes. Inspired by masked image modeling pre-training, we also devise efficient masking strategies for pre-training each of the spectral and spatial transformers. We conduct experiments on six publicly available datasets for HSI classification task and demonstrate that our model achieves state-of-the-art performance in all the datasets. The code for our model will be made available at https://github.com/csiro-robotics/factoformer.
Authors: Fabian Isensee, Klaus H.Maier-Hein
We participate in the AutoPET II challenge by modifying nnU-Net only through its easy to understand and modify 'nnUNetPlans.json' file. By switching to a UNet with residual encoder, increasing the batch size and increasing the patch size we obtain a configuration that substantially outperforms the automatically configured nnU-Net baseline (5-fold cross-validation Dice score of 65.14 vs 33.28) at the expense of increased compute requirements for model training. Our final submission ensembles the two most promising configurations.
Authors: Liu Jun, Zhou Jiantao, Zeng Jiandian, Jinyu Tian
This work investigates efficient score-based black-box adversarial attacks with a high Attack Success Rate (ASR) and good generalizability. We design a novel attack method based on a Disentangled Feature space, called DifAttack, which differs significantly from the existing ones operating over the entire feature space. Specifically, DifAttack firstly disentangles an image's latent feature into an adversarial feature and a visual feature, where the former dominates the adversarial capability of an image, while the latter largely determines its visual appearance. We train an autoencoder for the disentanglement by using pairs of clean images and their Adversarial Examples (AEs) generated from available surrogate models via white-box attack methods. Eventually, DifAttack iteratively optimizes the adversarial feature according to the query feedback from the victim model until a successful AE is generated, while keeping the visual feature unaltered. In addition, due to the avoidance of using surrogate models' gradient information when optimizing AEs for black-box models, our proposed DifAttack inherently possesses better attack capability in the open-set scenario, where the training dataset of the victim model is unknown. Extensive experimental results demonstrate that our method achieves significant improvements in ASR and query efficiency simultaneously, especially in the targeted attack and open-set scenarios. The code will be available at https://github.com/csjunjun/DifAttack.git soon.
Authors: Ethan Brewer, Giovani Valdrighi, Parikshit Solunke, Joao Rulff, Yurii Piadyk, Zhonghui Lv, Jorge Poco, Claudio Silva
Many areas of the world are without basic information on the socioeconomic well-being of the residing population due to limitations in existing data collection methods. Overhead images obtained remotely, such as from satellite or aircraft, can help serve as windows into the state of life on the ground and help "fill in the gaps" where community information is sparse, with estimates at smaller geographic scales requiring higher resolution sensors. Concurrent with improved sensor resolutions, recent advancements in machine learning and computer vision have made it possible to quickly extract features from and detect patterns in image data, in the process correlating these features with other information. In this work, we explore how well two approaches, a supervised convolutional neural network and semi-supervised clustering based on bag-of-visual-words, estimate population density, median household income, and educational attainment of individual neighborhoods from publicly available high-resolution imagery of cities throughout the United States. Results and analyses indicate that features extracted from the imagery can accurately estimate the density (R$^2$ up to 0.81) of neighborhoods, with the supervised approach able to explain about half the variation in a population's income and education. In addition to the presented approaches serving as a basis for further geographic generalization, the novel semi-supervised approach provides a foundation for future work seeking to estimate fine-scale information from aerial imagery without the need for label data.
Authors: Giacomo Aldegheri, Alina Rogalska, Ahmed Youssef, Eugenia Iofinova
In this work, we propose a method to 'hack' generative models, pushing their outputs away from the original training distribution towards a new objective. We inject a small-scale trainable module between the intermediate layers of the model and train it for a low number of iterations, keeping the rest of the network frozen. The resulting output images display an uncanny quality, given by the tension between the original and new objectives that can be exploited for artistic purposes.
Authors: Qinglong Cao, Zhengqin Xu, Yuntian Chen, Chao Ma, Xiaokang Yang
Large pre-trained vision-language models, such as CLIP, have shown remarkable generalization capabilities across various tasks when appropriate text prompts are provided. However, adapting these models to specific domains, like remote sensing images (RSIs), medical images, etc, remains unexplored and challenging. Existing prompt learning methods often lack domain-awareness or domain-transfer mechanisms, leading to suboptimal performance due to the misinterpretation of specific images in natural image patterns. To tackle this dilemma, we proposed a \textbf{Domain-Controlled Prompt Learning} for the specific domains. Specifically, the large-scale specific domain foundation model (LSDM) is first introduced to provide essential specific domain knowledge. Using lightweight neural networks, we transfer this knowledge into domain biases, which control both the visual and language branches to obtain domain-adaptive prompts in a directly incorporating manner. Simultaneously, to overcome the existing overfitting challenge, we propose a novel noisy-adding strategy, without extra trainable parameters, to help the model escape the suboptimal solution in a global domain oscillation manner. Experimental results show our method achieves state-of-the-art performance in specific domain image recognition datasets. Our code is available at https://github.com/caoql98/DCPL.
Authors: Zijie Pan, Jiachen Lu, Xiatian Zhu, Li Zhang
High-resolution 3D object generation remains a challenging task primarily due to the limited availability of comprehensive annotated training data. Recent advancements have aimed to overcome this constraint by harnessing image generative models, pretrained on extensive curated web datasets, using knowledge transfer techniques like Score Distillation Sampling (SDS). Efficiently addressing the requirements of high-resolution rendering often necessitates the adoption of latent representation-based models, such as the Latent Diffusion Model (LDM). In this framework, a significant challenge arises: To compute gradients for individual image pixels, it is necessary to backpropagate gradients from the designated latent space through the frozen components of the image model, such as the VAE encoder used within LDM. However, this gradient propagation pathway has never been optimized, remaining uncontrolled during training. We find that the unregulated gradients adversely affect the 3D model's capacity in acquiring texture-related information from the image generative model, leading to poor quality appearance synthesis. To address this overarching challenge, we propose an innovative operation termed Pixel-wise Gradient Clipping (PGC) designed for seamless integration into existing 3D generative models, thereby enhancing their synthesis quality. Specifically, we control the magnitude of stochastic gradients by clipping the pixel-wise gradients efficiently, while preserving crucial texture-related gradient directions. Despite this simplicity and minimal extra cost, extensive experiments demonstrate the efficacy of our PGC in enhancing the performance of existing 3D generative models for high-resolution object rendering.
Authors: Weixi Weng, Chun Yuan
Unsupervised domain adaptation object detection(UDAOD) research on Detection Transformer(DETR) mainly focuses on feature alignment and existing methods can be divided into two kinds, each of which has its unresolved issues. One-stage feature alignment methods can easily lead to performance fluctuation and training stagnation. Two-stage feature alignment method based on mean teacher comprises a pretraining stage followed by a self-training stage, each facing problems in obtaining reliable pretrained model and achieving consistent performance gains. Methods mentioned above have not yet explore how to utilize the third related domain such as target-like domain to assist adaptation. To address these issues, we propose a two-stage framework named MTM, i.e. Mean Teacher-DETR with Masked Feature Alignment. In the pretraining stage, we utilize labeled target-like images produced by image style transfer to avoid performance fluctuation. In the self-training stage, we leverage unlabeled target images by pseudo labels based on mean teacher and propose a module called Object Queries Knowledge Transfer(OQKT) to ensure consistent performance gains of the student model. Most importantly, we propose masked feature alignment methods including Masked Domain Query-based Feature Alignment(MDQFA) and Masked Token-wise Feature Alignment(MTWFA) to alleviate domain shift in a more robust way, which not only prevent training stagnation and lead to a robust pretrained model in the pretraining stage, but also enhance the model's target performance in the self-training stage. Experiments on three challenging scenarios and a theoretical analysis verify the effectiveness of MTM.
Authors: Aditya Kannan, Kenneth Shaw, Shikhar Bahl, Pragna Mannam, Deepak Pathak
Dexterity is often seen as a cornerstone of complex manipulation. Humans are able to perform a host of skills with their hands, from making food to operating tools. In this paper, we investigate these challenges, especially in the case of soft, deformable objects as well as complex, relatively long-horizon tasks. However, learning such behaviors from scratch can be data inefficient. To circumvent this, we propose a novel approach, DEFT (DExterous Fine-Tuning for Hand Policies), that leverages human-driven priors, which are executed directly in the real world. In order to improve upon these priors, DEFT involves an efficient online optimization procedure. With the integration of human-based learning and online fine-tuning, coupled with a soft robotic hand, DEFT demonstrates success across various tasks, establishing a robust, data-efficient pathway toward general dexterous manipulation. Please see our website at https://dexterous-finetuning.github.io for video results.
Authors: Giuseppe Stragapede, Ruben Vera-Rodriguez, Ruben Tolosana, Aythami Morales, Naser Damer, Julian Fierrez, Javier Ortega-Garcia
Analyzing keystroke dynamics (KD) for biometric verification has several advantages: it is among the most discriminative behavioral traits; keyboards are among the most common human-computer interfaces, being the primary means for users to enter textual data; its acquisition does not require additional hardware, and its processing is relatively lightweight; and it allows for transparently recognizing subjects. However, the heterogeneity of experimental protocols and metrics, and the limited size of the databases adopted in the literature impede direct comparisons between different systems, thus representing an obstacle in the advancement of keystroke biometrics. To alleviate this aspect, we present a new experimental framework to benchmark KD-based biometric verification performance and fairness based on tweet-long sequences of variable transcript text from over 185,000 subjects, acquired through desktop and mobile keyboards, extracted from the Aalto Keystroke Databases. The framework runs on CodaLab in the form of the Keystroke Verification Challenge (KVC). Moreover, we also introduce a novel fairness metric, the Skewed Impostor Ratio (SIR), to capture inter- and intra-demographic group bias patterns in the verification scores. We demonstrate the usefulness of the proposed framework by employing two state-of-the-art keystroke verification systems, TypeNet and TypeFormer, to compare different sets of input features, achieving a less privacy-invasive system, by discarding the analysis of text content (ASCII codes of the keys pressed) in favor of extended features in the time domain. Our experiments show that this approach allows to maintain satisfactory performance.
Authors: Gabriel De Araujo, Shanlin Sun, Xiaohui Xie
Image registration has traditionally been done using two distinct approaches: learning based methods, relying on robust deep neural networks, and optimization-based methods, applying complex mathematical transformations to warp images accordingly. Of course, both paradigms offer advantages and disadvantages, and, in this work, we seek to combine their respective strengths into a single streamlined framework, using the outputs of the learning based method as initial parameters for optimization while prioritizing computational power for the image pairs that offer the greatest loss. Our investigations showed that an improvement of 1.5% in testing when utilizing the best performing state-of-the-art model as the backbone of the framework, while maintaining the same inference time and a substantial 0.94% points performance gain in deformation field smoothness.
Authors: Bin Xia, Shiyin Wang, Yingfan Tao, Yitong Wang, Jiaya Jia
In this paper, we introduce a Multimodal Large Language Model-based Generation Assistant (LLMGA), leveraging the vast reservoir of knowledge and proficiency in reasoning, comprehension, and response inherent in Large Language Models (LLMs) to assist users in image generation and editing. Diverging from existing approaches where Multimodal Large Language Models (MLLMs) generate fixed-size embeddings to control Stable Diffusion (SD), our LLMGA provides a detailed language generation prompt for precise control over SD. This not only augments LLM context understanding but also reduces noise in generation prompts, yields images with more intricate and precise content, and elevates the interpretability of the network. To this end, we curate a comprehensive dataset comprising prompt refinement, similar image generation, inpainting $\&$ outpainting, and visual question answering. Moreover, we propose a two-stage training scheme. In the first stage, we train the MLLM to grasp the properties of image generation and editing, enabling it to generate detailed prompts. In the second stage, we optimize SD to align with the MLLM's generation prompts. Additionally, we propose a reference-based restoration network to alleviate texture, brightness, and contrast disparities between generated and preserved regions during image editing. Extensive results show that LLMGA has promising generative capabilities and can enable wider applications in an interactive manner.
Authors: Wenjie Zhao, Jia Li, Xin Dong, Yu Xiang, Yunhui Guo
Semantic segmentation models, while effective for in-distribution categories, face challenges in real-world deployment due to encountering out-of-distribution (OoD) objects. Detecting these OoD objects is crucial for safety-critical applications. Existing methods rely on anomaly scores, but choosing a suitable threshold for generating masks presents difficulties and can lead to fragmentation and inaccuracy. This paper introduces a method to convert anomaly \underline{S}core \underline{T}o segmentation \underline{M}ask, called S2M, a simple and effective framework for OoD detection in semantic segmentation. Unlike assigning anomaly scores to pixels, S2M directly segments the entire OoD object. By transforming anomaly scores into prompts for a promptable segmentation model, S2M eliminates the need for threshold selection. Extensive experiments demonstrate that S2M outperforms the state-of-the-art by approximately 10% in IoU and 30% in mean F1 score, on average, across various benchmarks including Fishyscapes, Segment-Me-If-You-Can, and RoadAnomaly datasets.
Authors: Wooseok Kim, Taiki Fukiage, Takeshi Oishi
Recently, significant progress has been made in the study of methods for 3D reconstruction from multiple images using implicit neural representations, exemplified by the neural radiance field (NeRF) method. Such methods, which are based on volume rendering, can model various light phenomena, and various extended methods have been proposed to accommodate different scenes and situations. However, when handling scenes with multiple glass objects, e.g., objects in a glass showcase, modeling the target scene accurately has been challenging due to the presence of multiple reflection and refraction effects. Thus, this paper proposes a NeRF-based modeling method for scenes containing a glass case. In the proposed method, refraction and reflection are modeled using elements that are dependent and independent of the viewer's perspective. This approach allows us to estimate the surfaces where refraction occurs, i.e., glass surfaces, and enables the separation and modeling of both direct and reflected light components. Compared to existing methods, the proposed method enables more accurate modeling of both glass refraction and the overall scene.
Authors: Thomas Dooms, Ing Jyh Tsang, Jose Oramas
Modern machine learning models are able to outperform humans on a variety of non-trivial tasks. However, as the complexity of the models increases, they consume significant amounts of power and still struggle to generalize effectively to unseen data. Local learning, which focuses on updating subsets of a model's parameters at a time, has emerged as a promising technique to address these issues. Recently, a novel local learning algorithm, called Forward-Forward, has received widespread attention due to its innovative approach to learning. Unfortunately, its application has been limited to smaller datasets due to scalability issues. To this end, we propose The Trifecta, a collection of three simple techniques that synergize exceptionally well and drastically improve the Forward-Forward algorithm on deeper networks. Our experiments demonstrate that our models are on par with similarly structured, backpropagation-based models in both training speed and test accuracy on simple datasets. This is achieved by the ability to learn representations that are informative locally, on a layer-by-layer basis, and retain their informativeness when propagated to deeper layers in the architecture. This leads to around 84% accuracy on CIFAR-10, a notable improvement (25%) over the original FF algorithm. These results highlight the potential of Forward-Forward as a genuine competitor to backpropagation and as a promising research avenue.
Authors: Xianda Guo, Juntao Lu, Chenming Zhang, Yiqi Wang, Yiqun Duan, Tian Yang, Zheng Zhu, Long Chen
Stereo matching, a pivotal technique in computer vision, plays a crucial role in robotics, autonomous navigation, and augmented reality. Despite the development of numerous impressive methods in recent years, replicating their results and determining the most suitable architecture for practical application remains challenging. Addressing this gap, our paper introduces a comprehensive benchmark focusing on practical applicability rather than solely on performance enhancement. Specifically, we develop a flexible and efficient stereo matching codebase, called OpenStereo. OpenStereo includes training and inference codes of more than 12 network models, making it, to our knowledge, the most complete stereo matching toolbox available. Based on OpenStereo, we conducted experiments on the SceneFlow dataset and have achieved or surpassed the performance metrics reported in the original paper. Additionally, we conduct an in-depth revisitation of recent developments in stereo matching through ablative experiments. These investigations inspired the creation of StereoBase, a simple yet strong baseline model. Our extensive comparative analyses of StereoBase against numerous contemporary stereo matching methods on the SceneFlow dataset demonstrate its remarkably strong performance. The source code is available at https://github.com/XiandaGuo/OpenStereo.
Authors: Yizhi Wang, Wallace Lira, Wenqi Wang, Ali Mahdavi-Amiri, Hao Zhang
We introduce multi-slice reasoning, a new notion for single-view 3D reconstruction which challenges the current and prevailing belief that multi-view synthesis is the most natural conduit between single-view and 3D. Our key observation is that object slicing is more advantageous than altering views to reveal occluded structures. Specifically, slicing is more occlusion-revealing since it can peel through any occluders without obstruction. In the limit, i.e., with infinitely many slices, it is guaranteed to unveil all hidden object parts. We realize our idea by developing Slice3D, a novel method for single-view 3D reconstruction which first predicts multi-slice images from a single RGB image and then integrates the slices into a 3D model using a coordinate-based transformer network for signed distance prediction. The slice images can be regressed or generated, both through a U-Net based network. For the former, we inject a learnable slice indicator code to designate each decoded image into a spatial slice location, while the slice generator is a denoising diffusion model operating on the entirety of slice images stacked on the input channels. We conduct extensive evaluation against state-of-the-art alternatives to demonstrate superiority of our method, especially in recovering complex and severely occluded shape structures, amid ambiguities. All Slice3D results were produced by networks trained on a single Nvidia A40 GPU, with an inference time less than 20 seconds.
Authors: Cong Wang, Jiaxi Gu, Panwen Hu, Songcen Xu, Hang Xu, Xiaodan Liang
Image-to-video generation, which aims to generate a video starting from a given reference image, has drawn great attention. Existing methods try to extend pre-trained text-guided image diffusion models to image-guided video generation models. Nevertheless, these methods often result in either low fidelity or flickering over time due to their limitation to shallow image guidance and poor temporal consistency. To tackle these problems, we propose a high-fidelity image-to-video generation method by devising a frame retention branch based on a pre-trained video diffusion model, named DreamVideo. Instead of integrating the reference image into the diffusion process at a semantic level, our DreamVideo perceives the reference image via convolution layers and concatenates the features with the noisy latents as model input. By this means, the details of the reference image can be preserved to the greatest extent. In addition, by incorporating double-condition classifier-free guidance, a single image can be directed to videos of different actions by providing varying prompt texts. This has significant implications for controllable video generation and holds broad application prospects. We conduct comprehensive experiments on the public dataset, and both quantitative and qualitative results indicate that our method outperforms the state-of-the-art method. Especially for fidelity, our model has a powerful image retention ability and delivers the best results in UCF101 compared to other image-to-video models to our best knowledge. Also, precise control can be achieved by giving different text prompts. Further details and comprehensive results of our model will be presented in https://anonymous0769.github.io/DreamVideo/.
Authors: Wenxuan Guo, Zhiyu Pan, Yingping Liang, Ziheng Xi, Zhi Chen Zhong, Jianjiang Feng, Jie Zhou
Camera-based person re-identification (ReID) systems have been widely applied in the field of public security. However, cameras often lack the perception of 3D morphological information of human and are susceptible to various limitations, such as inadequate illumination, complex background, and personal privacy. In this paper, we propose a LiDAR-based ReID framework, ReID3D, that utilizes pre-training strategy to retrieve features of 3D body shape and introduces Graph-based Complementary Enhancement Encoder for extracting comprehensive features. Due to the lack of LiDAR datasets, we build LReID, the first LiDAR-based person ReID dataset, which is collected in several outdoor scenes with variations in natural conditions. Additionally, we introduce LReID-sync, a simulated pedestrian dataset designed for pre-training encoders with tasks of point cloud completion and shape parameter learning. Extensive experiments on LReID show that ReID3D achieves exceptional performance with a rank-1 accuracy of 94.0, highlighting the significant potential of LiDAR in addressing person ReID tasks. To the best of our knowledge, we are the first to propose a solution for LiDAR-based ReID. The code and datasets will be released soon.
Authors: Junhao Zhuang, Yanhong Zeng, Wenran Liu, Chun Yuan, Kai Chen
Achieving high-quality versatile image inpainting, where user-specified regions are filled with plausible content according to user intent, presents a significant challenge. Existing methods face difficulties in simultaneously addressing context-aware image inpainting and text-guided object inpainting due to the distinct optimal training strategies required. To overcome this challenge, we introduce PowerPaint, the first high-quality and versatile inpainting model that excels in both tasks. First, we introduce learnable task prompts along with tailored fine-tuning strategies to guide the model's focus on different inpainting targets explicitly. This enables PowerPaint to accomplish various inpainting tasks by utilizing different task prompts, resulting in state-of-the-art performance. Second, we demonstrate the versatility of the task prompt in PowerPaint by showcasing its effectiveness as a negative prompt for object removal. Additionally, we leverage prompt interpolation techniques to enable controllable shape-guided object inpainting. Finally, we extensively evaluate PowerPaint on various inpainting benchmarks to demonstrate its superior performance for versatile image inpainting. We release our codes and models on our project page: https://powerpaint.github.io/.
Authors: Yuhang Lu, Xinge Zhu, Tai Wang, Yuexin Ma
Occupancy prediction has increasingly garnered attention in recent years for its fine-grained understanding of 3D scenes. Traditional approaches typically rely on dense, regular grid representations, which often leads to excessive computational demands and a loss of spatial details for small objects. This paper introduces OctreeOcc, an innovative 3D occupancy prediction framework that leverages the octree representation to adaptively capture valuable information in 3D, offering variable granularity to accommodate object shapes and semantic regions of varying sizes and complexities. In particular, we incorporate image semantic information to improve the accuracy of initial octree structures and design an effective rectification mechanism to refine the octree structure iteratively. Our extensive evaluations show that OctreeOcc not only surpasses state-of-the-art methods in occupancy prediction, but also achieves a 15%-24% reduction in computational overhead compared to dense-grid-based methods.
Authors: Xin Li, Yeqi Bai, Pinlong Cai, Licheng Wen, Daocheng Fu, Bo Zhang, Xuemeng Yang, Xinyu Cai, Tao Ma, Jianfei Guo, Xing Gao, Min Dou, Botian Shi, Yong Liu, Liang He, Yu Qiao
This paper explores the emerging knowledge-driven autonomous driving technologies. Our investigation highlights the limitations of current autonomous driving systems, in particular their sensitivity to data bias, difficulty in handling long-tail scenarios, and lack of interpretability. Conversely, knowledge-driven methods with the abilities of cognition, generalization and life-long learning emerge as a promising way to overcome these challenges. This paper delves into the essence of knowledge-driven autonomous driving and examines its core components: dataset \& benchmark, environment, and driver agent. By leveraging large language models, world models, neural rendering, and other advanced artificial intelligence techniques, these components collectively contribute to a more holistic, adaptive, and intelligent autonomous driving system. The paper systematically organizes and reviews previous research efforts in this area, and provides insights and guidance for future research and practical applications of autonomous driving. We will continually share the latest updates on cutting-edge developments in knowledge-driven autonomous driving along with the relevant valuable open-source resources at: \url{https://github.com/PJLab-ADG/awesome-knowledge-driven-AD}.
Authors: Pengcheng Chen, Ziyan Huang, Zhongying Deng, Tianbin Li, Yanzhou Su, Haoyu Wang, Jin Ye, Yu Qiao, Junjun He
OpenAI's latest large vision-language model (LVLM), GPT-4V(ision), has piqued considerable interest for its potential in medical applications. Despite its promise, recent studies and internal reviews highlight its underperformance in specialized medical tasks. This paper explores the boundary of GPT-4V's capabilities in medicine, particularly in processing complex imaging data from endoscopies, CT scans, and MRIs etc. Leveraging open-source datasets, we assessed its foundational competencies, identifying substantial areas for enhancement. Our research emphasizes prompt engineering, an often-underutilized strategy for improving AI responsiveness. Through iterative testing, we refined the model's prompts, significantly improving its interpretative accuracy and relevance in medical imaging. From our comprehensive evaluations, we distilled 10 effective prompt engineering techniques, each fortifying GPT-4V's medical acumen. These methodical enhancements facilitate more reliable, precise, and clinically valuable insights from GPT-4V, advancing its operability in critical healthcare environments. Our findings are pivotal for those employing AI in medicine, providing clear, actionable guidance on harnessing GPT-4V's full diagnostic potential.
Authors: Jae Sung Park, Jack Hessel, Khyathi Raghavi Chandu, Paul Pu Liang, Ximing Lu, Peter West, Youngjae Yu, Qiuyuan Huang, Jianfeng Gao, Ali Farhadi, Yejin Choi
Instruction following vision-language (VL) models offer a flexible interface that supports a broad range of multimodal tasks in a zero-shot fashion. However, interfaces that operate on full images do not directly enable the user to "point to" and access specific regions within images. This capability is important not only to support reference-grounded VL benchmarks, but also, for practical applications that require precise within-image reasoning. We build Localized Visual Commonsense models, which allow users to specify (multiple) regions as input. We train our model by sampling localized commonsense knowledge from a large language model (LLM): specifically, we prompt an LLM to collect commonsense knowledge given a global literal image description and a local literal region description automatically generated by a set of VL models. With a separately trained critic model that selects high-quality examples, we find that training on the localized commonsense corpus can successfully distill existing VL models to support a reference-as-input interface. Empirical results and human evaluations in a zero-shot setup demonstrate that our distillation method results in more precise VL models of reasoning compared to a baseline of passing a generated referring expression to an LLM.
Authors: Tianjin Huang, Tianlong Chen, Zhangyang Wang, Shiwei Liu
Vision Transformers have been rapidly uprising in computer vision thanks to their outstanding scaling trends, and gradually replacing convolutional neural networks (CNNs). Recent works on self-supervised learning (SSL) introduce siamese pre-training tasks, on which Transformer backbones continue to demonstrate ever stronger results than CNNs. People come to believe that Transformers or self-attention modules are inherently more suitable than CNNs in the context of SSL. However, it is noteworthy that most if not all prior arts of SSL with CNNs chose the standard ResNets as their backbones, whose architecture effectiveness is known to already lag behind advanced Vision Transformers. Therefore, it remains unclear whether the self-attention operation is crucial for the recent advances in SSL - or CNNs can deliver the same excellence with more advanced designs, too? Can we close the SSL performance gap between Transformers and CNNs? To answer these intriguing questions, we apply self-supervised pre-training to the recently proposed, stronger lager-kernel CNN architecture and conduct an apple-to-apple comparison with Transformers, in their SSL performance. Our results show that we are able to build pure CNN SSL architectures that perform on par with or better than the best SSL-trained Transformers, by just scaling up convolutional kernel sizes besides other small tweaks. Impressively, when transferring to the downstream tasks \texttt{MS COCO} detection and segmentation, our SSL pre-trained CNN model (trained in 100 epochs) achieves the same good performance as the 300-epoch pre-trained Transformer counterpart. We hope this work can help to better understand what is essential (or not) for self-supervised learning backbones.
Authors: Boyu Shi, Shiyu Xia, Xu Yang, Haokun Chen, Zhiqiang Kou, Xin Geng
Recently, Stitchable Neural Networks (SN-Net) is proposed to stitch some pre-trained networks for quickly building numerous networks with different complexity and performance trade-offs. In this way, the burdens of designing or training the variable-sized networks, which can be used in application scenarios with diverse resource constraints, are alleviated. However, SN-Net still faces a few challenges. 1) Stitching from multiple independently pre-trained anchors introduces high storage resource consumption. 2) SN-Net faces challenges to build smaller models for low resource constraints. 3). SN-Net uses an unlearned initialization method for stitch layers, limiting the final performance. To overcome these challenges, motivated by the recently proposed Learngene framework, we propose a novel method called Learngene Pool. Briefly, Learngene distills the critical knowledge from a large pre-trained model into a small part (termed as learngene) and then expands this small part into a few variable-sized models. In our proposed method, we distill one pretrained large model into multiple small models whose network blocks are used as learngene instances to construct the learngene pool. Since only one large model is used, we do not need to store more large models as SN-Net and after distilling, smaller learngene instances can be created to build small models to satisfy low resource constraints. We also insert learnable transformation matrices between the instances to stitch them into variable-sized models to improve the performance of these models. Exhaustive experiments have been implemented and the results validate the effectiveness of the proposed Learngene Pool compared with SN-Net.
Authors: Zhengxue Wang, Zhiqiang Yan, Jian Yang
Depth super-resolution (DSR) aims to restore high-resolution (HR) depth from low-resolution (LR) one, where RGB image is often used to promote this task. Recent image guided DSR approaches mainly focus on spatial domain to rebuild depth structure. However, since the structure of LR depth is usually blurry, only considering spatial domain is not very sufficient to acquire satisfactory results. In this paper, we propose structure guided network (SGNet), a method that pays more attention to gradient and frequency domains, both of which have the inherent ability to capture high-frequency structure. Specifically, we first introduce the gradient calibration module (GCM), which employs the accurate gradient prior of RGB to sharpen the LR depth structure. Then we present the Frequency Awareness Module (FAM) that recursively conducts multiple spectrum differencing blocks (SDB), each of which propagates the precise high-frequency components of RGB into the LR depth. Extensive experimental results on both real and synthetic datasets demonstrate the superiority of our SGNet, reaching the state-of-the-art. Codes and pre-trained models are available at https://github.com/yanzq95/SGNet.
Authors: Xiaojian Yuan, Kejiang Chen, Wen Huang, Jie Zhang, Weiming Zhang, Nenghai Yu
The popularity of Machine Learning as a Service (MLaaS) has led to increased concerns about Model Stealing Attacks (MSA), which aim to craft a clone model by querying MLaaS. Currently, most research on MSA assumes that MLaaS can provide soft labels and that the attacker has a proxy dataset with a similar distribution. However, this fails to encapsulate the more practical scenario where only hard labels are returned by MLaaS and the data distribution remains elusive. Furthermore, most existing work focuses solely on stealing the model accuracy, neglecting the model robustness, while robustness is essential in security-sensitive scenarios, e.g., face-scan payment. Notably, improving model robustness often necessitates the use of expensive techniques such as adversarial training, thereby further making stealing robustness a more lucrative prospect. In response to these identified gaps, we introduce a novel Data-Free Hard-Label Robustness Stealing (DFHL-RS) attack in this paper, which enables the stealing of both model accuracy and robustness by simply querying hard labels of the target model without the help of any natural data. Comprehensive experiments demonstrate the effectiveness of our method. The clone model achieves a clean accuracy of 77.86% and a robust accuracy of 39.51% against AutoAttack, which are only 4.71% and 8.40% lower than the target model on the CIFAR-10 dataset, significantly exceeding the baselines. Our code is available at: https://github.com/LetheSec/DFHL-RS-Attack.
Authors: Zihao Zhao, Sheng Wang, Qian Wang, Dinggang Shen
Obtaining large-scale radiology reports can be difficult for medical images due to various reasons, limiting the effectiveness of contrastive pre-training in the medical image domain and underscoring the need for alternative methods. In this paper, we propose eye-tracking as an alternative to text reports, as it allows for the passive collection of gaze signals without disturbing radiologist's routine diagnosis process. By tracking the gaze of radiologists as they read and diagnose medical images, we can understand their visual attention and clinical reasoning. When a radiologist has similar gazes for two medical images, it may indicate semantic similarity for diagnosis, and these images should be treated as positive pairs when pre-training a computer-assisted diagnosis (CAD) network through contrastive learning. Accordingly, we introduce the Medical contrastive Gaze Image Pre-training (McGIP) as a plug-and-play module for contrastive learning frameworks. McGIP uses radiologist's gaze to guide contrastive pre-training. We evaluate our method using two representative types of medical images and two common types of gaze data. The experimental results demonstrate the practicality of McGIP, indicating its high potential for various clinical scenarios and applications.
Authors: Zhiyu Pan, Zhicheng Zhong, Wenxuan Guo, Yifan Chen, Jianjiang Feng, Jie Zhou
Recently, several methods have been proposed to estimate 3D human pose from multi-view images and achieved impressive performance on public datasets collected in relatively easy scenarios. However, there are limited approaches for extracting 3D human skeletons from multimodal inputs (e.g., RGB and pointcloud) that can enhance the accuracy of predicting 3D poses in challenging situations. We fill this gap by introducing a pipeline called PointVoxel that fuses multi-view RGB and pointcloud inputs to obtain 3D human poses. We demonstrate that volumetric representation is an effective architecture for integrating these different modalities. Moreover, in order to overcome the challenges of annotating 3D human pose labels in difficult scenarios, we develop a synthetic dataset generator for pretraining and design an unsupervised domain adaptation strategy so that we can obtain a well-trained 3D human pose estimator without using any manual annotations. We evaluate our approach on four datasets (two public datasets, one synthetic dataset, and one challenging dataset named BasketBall collected by ourselves), showing promising results. The code and dataset will be released soon.
Authors: Rongkun Zheng, Lu Qi, Xi Chen, Yi Wang, Kun Wang, Yu Qiao, Hengshuang Zhao
Training on large-scale datasets can boost the performance of video instance segmentation while the annotated datasets for VIS are hard to scale up due to the high labor cost. What we possess are numerous isolated filed-specific datasets, thus, it is appealing to jointly train models across the aggregation of datasets to enhance data volume and diversity. However, due to the heterogeneity in category space, as mask precision increases with the data volume, simply utilizing multiple datasets will dilute the attention of models on different taxonomies. Thus, increasing the data scale and enriching taxonomy space while improving classification precision is important. In this work, we analyze that providing extra taxonomy information can help models concentrate on specific taxonomy, and propose our model named Taxonomy-aware Multi-dataset Joint Training for Video Instance Segmentation (TMT-VIS) to address this vital challenge. Specifically, we design a two-stage taxonomy aggregation module that first compiles taxonomy information from input videos and then aggregates these taxonomy priors into instance queries before the transformer decoder. We conduct extensive experimental evaluations on four popular and challenging benchmarks, including YouTube-VIS 2019, YouTube-VIS 2021, OVIS, and UVO. Our model shows significant improvement over the baseline solutions, and sets new state-of-the-art records on all benchmarks. These appealing and encouraging results demonstrate the effectiveness and generality of our approach. The code is available at https://github.com/rkzheng99/TMT-VIS(https://github.com/rkzheng99/TMT-VIS)
Authors: Pooja Prajod, Matteo Lavit Nicora, Marta Mondellini, Giovanni Tauro, Rocco Vertechy, Matteo Malosio, Elisabeth André
Collaborative robots (cobots) are widely used in industrial applications, yet extensive research is still needed to enhance human-robot collaborations and operator experience. A potential approach to improve the collaboration experience involves adapting cobot behavior based on natural cues from the operator. Inspired by the literature on human-human interactions, we conducted a wizard-of-oz study to examine whether a gaze towards the cobot can serve as a trigger for initiating joint activities in collaborative sessions. In this study, 37 participants engaged in an assembly task while their gaze behavior was analyzed. We employ a gaze-based attention recognition model to identify when the participants look at the cobot. Our results indicate that in most cases (84.88\%), the joint activity is preceded by a gaze towards the cobot. Furthermore, during the entire assembly cycle, the participants tend to look at the cobot around the time of the joint activity. To the best of our knowledge, this is the first study to analyze the natural gaze behavior of participants working on a joint activity with a robot during a collaborative assembly task.
Authors: Zhongang Cai, Jianping Jiang, Zhongfei Qing, Xinying Guo, Mingyuan Zhang, Zhengyu Lin, Haiyi Mei, Chen Wei, Ruisi Wang, Wanqi Yin, Xiangyu Fan, Han Du, Liang Pan, Peng Gao, Zhitao Yang, Yang Gao, Jiaqi Li, Tianxiang Ren, Yukun Wei, Xiaogang Wang, Chen Change Loy, Lei Yang, Ziwei Liu
In this work, we present Digital Life Project, a framework utilizing language as the universal medium to build autonomous 3D characters, who are capable of engaging in social interactions and expressing with articulated body motions, thereby simulating life in a digital environment. Our framework comprises two primary components: 1) SocioMind: a meticulously crafted digital brain that models personalities with systematic few-shot exemplars, incorporates a reflection process based on psychology principles, and emulates autonomy by initiating dialogue topics; 2) MoMat-MoGen: a text-driven motion synthesis paradigm for controlling the character's digital body. It integrates motion matching, a proven industry technique to ensure motion quality, with cutting-edge advancements in motion generation for diversity. Extensive experiments demonstrate that each module achieves state-of-the-art performance in its respective domain. Collectively, they enable virtual characters to initiate and sustain dialogues autonomously, while evolving their socio-psychological states. Concurrently, these characters can perform contextually relevant bodily movements. Additionally, a motion captioning module further allows the virtual character to recognize and appropriately respond to human players' actions. Homepage: https://digital-life-project.com/