Authors: Prosenjit Chatterjee, Alex Yalchin, Joseph Shelton, Kaushik Roy, Xiaohong Yuan, Kossi D. Edoh
Biometric authentication is becoming more prevalent for secured authentication systems. However, the biometric substances can be deceived by the imposters in several ways. Among other imposter attacks, print attacks, mask attacks, and replay attacks fall under the presentation attack category. The bio-metric images, especially the iris and face, are vulnerable to different presentation attacks. This research applies deep learning approaches to mitigate presentation attacks in a biometric access control system. Our contribution in this paper is two-fold: First, we applied the wavelet transform to extract the features from the biometric images. Second, we modified the deep residual neural net and applied it to the spoof datasets in an attempt to detect the presentation attacks. This research applied the proposed approach to biometric spoof datasets, namely ATVS, CASIA two class, and CASIA cropped image sets. The datasets used in this research contain images that are captured in both a controlled and uncontrolled environment along with different resolutions and sizes. We obtained the best accuracy of 93% on the ATVS Iris datasets. For CASIA two class and CASIA cropped datasets, we achieved test accuracies of 91% and 82%, respectively.
Authors: Justin Spencer, Deborah Lawrence, Prosenjit Chatterjee, Kaushik Roy, Albert Esterline, Jung-Hee Kim
The use of biometrics to authenticate users and control access to secure areas has become extremely popular in recent years, and biometric access control systems are frequently used by both governments and private corporations. However, these systems may represent risks to security when deployed without considering the possibility of biometric presentation attacks (also known as spoofing). Presentation attacks are a serious threat because they do not require significant time, expense, or skill to carry out while remaining effective against many biometric systems in use today. This research compares three different software-based methods for facial and iris presentation attack detection in images. The first method uses Inception-v3, a pre-trained deep Convolutional Neural Network (CNN) made by Google for the ImageNet challenge, which is retrained for this problem. The second uses a shallow CNN based on a modified Spoofnet architecture, which is trained normally. The third is a texture-based method using Local Binary Patterns (LBP). The datasets used are the ATVS-FIr dataset, which contains real and fake iris images, and the CASIA Face Anti-Spoofing Dataset, which contains real images as well as warped photos, cut photos, and video replay presentation attacks. We also present a third set of results, based on cropped versions of the CASIA images.
Authors: Eadom Dessalene, Michael Maynord, Cornelia Fermüller, Yiannis Aloimonos
We introduce LEAP (illustrated in Figure 1), a novel method for generating video-grounded action programs through use of a Large Language Model (LLM). These action programs represent the motoric, perceptual, and structural aspects of action, and consist of sub-actions, pre- and post-conditions, and control flows. LEAP's action programs are centered on egocentric video and employ recent developments in LLMs both as a source for program knowledge and as an aggregator and assessor of multimodal video information. We apply LEAP over a majority (87\%) of the training set of the EPIC Kitchens dataset, and release the resulting action programs as a publicly available dataset here (https://drive.google.com/drive/folders/1Cpkw_TI1IIxXdzor0pOXG3rWJWuKU5Ex?usp=drive_link). We employ LEAP as a secondary source of supervision, using its action programs in a loss term applied to action recognition and anticipation networks. We demonstrate sizable improvements in performance in both tasks due to training with the LEAP dataset. Our method achieves 1st place on the EPIC Kitchens Action Recognition leaderboard as of November 17 among the networks restricted to RGB-input (see Supplementary Materials).
Authors: Xiang Li, Qianli Shen, Kenji Kawaguchi
The booming use of text-to-image generative models has raised concerns about their high risk of producing copyright-infringing content. While probabilistic copyright protection methods provide a probabilistic guarantee against such infringement, in this paper, we introduce Virtually Assured Amplification Attack (VA3), a novel online attack framework that exposes the vulnerabilities of these protection mechanisms. The proposed framework significantly amplifies the probability of generating infringing content on the sustained interactions with generative models and a lower-bounded success probability of each engagement. Our theoretical and experimental results demonstrate the effectiveness of our approach and highlight the potential risk of implementing probabilistic copyright protection in practical applications of text-to-image generative models. Code is available at https://github.com/South7X/VA3.
Authors: Chuan Guo, Yuxuan Mu, Muhammad Gohar Javed, Sen Wang, Li Cheng
We introduce MoMask, a novel masked modeling framework for text-driven 3D human motion generation. In MoMask, a hierarchical quantization scheme is employed to represent human motion as multi-layer discrete motion tokens with high-fidelity details. Starting at the base layer, with a sequence of motion tokens obtained by vector quantization, the residual tokens of increasing orders are derived and stored at the subsequent layers of the hierarchy. This is consequently followed by two distinct bidirectional transformers. For the base-layer motion tokens, a Masked Transformer is designated to predict randomly masked motion tokens conditioned on text input at training stage. During generation (i.e. inference) stage, starting from an empty sequence, our Masked Transformer iteratively fills up the missing tokens; Subsequently, a Residual Transformer learns to progressively predict the next-layer tokens based on the results from current layer. Extensive experiments demonstrate that MoMask outperforms the state-of-art methods on the text-to-motion generation task, with an FID of 0.045 (vs e.g. 0.141 of T2M-GPT) on the HumanML3D dataset, and 0.228 (vs 0.514) on KIT-ML, respectively. MoMask can also be seamlessly applied in related tasks without further model fine-tuning, such as text-guided temporal inpainting.
Authors: Eric Hedlin, Gopal Sharma, Shweta Mahajan, Xingzhe He, Hossam Isack, Abhishek Kar Helge Rhodin, Andrea Tagliasacchi, Kwang Moo Yi
Unsupervised learning of keypoints and landmarks has seen significant progress with the help of modern neural network architectures, but performance is yet to match the supervised counterpart, making their practicability questionable. We leverage the emergent knowledge within text-to-image diffusion models, towards more robust unsupervised keypoints. Our core idea is to find text embeddings that would cause the generative model to consistently attend to compact regions in images (i.e. keypoints). To do so, we simply optimize the text embedding such that the cross-attention maps within the denoising network are localized as Gaussians with small standard deviations. We validate our performance on multiple datasets: the CelebA, CUB-200-2011, Tai-Chi-HD, DeepFashion, and Human3.6m datasets. We achieve significantly improved accuracy, sometimes even outperforming supervised ones, particularly for data that is non-aligned and less curated. Our code is publicly available and can be found through our project page: https://ubc-vision.github.io/StableKeypoints/
Authors: Lukas Hirsch, Yu Huang, Hernan A. Makse, Danny F. Martinez, Mary Hughes, Sarah Eskreis-Winkler, Katja Pinker, Elizabeth Morris, Lucas C. Parra, Elizabeth J. Sutton
Women with an increased life-time risk of breast cancer undergo supplemental annual screening MRI. We propose to predict the risk of developing breast cancer within one year based on the current MRI, with the objective of reducing screening burden and facilitating early detection. An AI algorithm was developed on 53,858 breasts from 12,694 patients who underwent screening or diagnostic MRI and accrued over 12 years, with 2,331 confirmed cancers. A first U-Net was trained to segment lesions and identify regions of concern. A second convolutional network was trained to detect malignant cancer using features extracted by the U-Net. This network was then fine-tuned to estimate the risk of developing cancer within a year in cases that radiologists considered normal or likely benign. Risk predictions from this AI were evaluated with a retrospective analysis of 9,183 breasts from a high-risk screening cohort, which were not used for training. Statistical analysis focused on the tradeoff between number of omitted exams versus negative predictive value, and number of potential early detections versus positive predictive value. The AI algorithm identified regions of concern that coincided with future tumors in 52% of screen-detected cancers. Upon directed review, a radiologist found that 71.3% of cancers had a visible correlate on the MRI prior to diagnosis, 65% of these correlates were identified by the AI model. Reevaluating these regions in 10% of all cases with higher AI-predicted risk could have resulted in up to 33% early detections by a radiologist. Additionally, screening burden could have been reduced in 16% of lower-risk cases by recommending a later follow-up without compromising current interval cancer rate. With increasing datasets and improving image quality we expect this new AI-aided, adaptive screening to meaningfully reduce screening burden and improve early detection.
Authors: Prashant Kumar, Kshitij Madhav Bhat, Vedang Bhupesh Shenvi Nadkarni, Prem Kalra
Sparse LiDAR point clouds cause severe loss of detail of static structures and reduce the density of static points available for navigation. Reduced density can be detrimental to navigation under several scenarios. We observe that despite high sparsity, in most cases, the global topology of LiDAR outlining the static structures can be inferred. We utilize this property to obtain a backbone skeleton of a static LiDAR scan in the form of a single connected component that is a proxy to its global topology. We utilize the backbone to augment new points along static structures to overcome sparsity. Newly introduced points could correspond to existing static structures or to static points that were earlier obstructed by dynamic objects. To the best of our knowledge, we are the first to use this strategy for sparse LiDAR point clouds. Existing solutions close to our approach fail to identify and preserve the global static LiDAR topology and generate sub-optimal points. We propose GLiDR, a Graph Generative network that is topologically regularized using 0-dimensional Persistent Homology (PH) constraints. This enables GLiDR to introduce newer static points along a topologically consistent global static LiDAR backbone. GLiDR generates precise static points using 32x sparser dynamic scans and performs better than the baselines across three datasets. The newly introduced static points allow GLiDR to outperform LiDAR-based navigation using SLAM in several settings. GLiDR generates a valuable byproduct - an accurate binary segmentation mask of static and dynamic objects that is helpful for navigation planning and safety in constrained environments.
Authors: Depanshu Sani, Sandeep Mahato, Sourabh Saini, Harsh Kumar Agarwal, Charu Chandra Devshali, Saket Anand, Gaurav Arora, Thiagarajan Jayaraman
The availability of well-curated datasets has driven the success of Machine Learning (ML) models. Despite greater access to earth observation data in agriculture, there is a scarcity of curated and labelled datasets, which limits the potential of its use in training ML models for remote sensing (RS) in agriculture. To this end, we introduce a first-of-its-kind dataset called SICKLE, which constitutes a time-series of multi-resolution imagery from 3 distinct satellites: Landsat-8, Sentinel-1 and Sentinel-2. Our dataset constitutes multi-spectral, thermal and microwave sensors during January 2018 - March 2021 period. We construct each temporal sequence by considering the cropping practices followed by farmers primarily engaged in paddy cultivation in the Cauvery Delta region of Tamil Nadu, India; and annotate the corresponding imagery with key cropping parameters at multiple resolutions (i.e. 3m, 10m and 30m). Our dataset comprises 2,370 season-wise samples from 388 unique plots, having an average size of 0.38 acres, for classifying 21 crop types across 4 districts in the Delta, which amounts to approximately 209,000 satellite images. Out of the 2,370 samples, 351 paddy samples from 145 plots are annotated with multiple crop parameters; such as the variety of paddy, its growing season and productivity in terms of per-acre yields. Ours is also one among the first studies that consider the growing season activities pertinent to crop phenology (spans sowing, transplanting and harvesting dates) as parameters of interest. We benchmark SICKLE on three tasks: crop type, crop phenology (sowing, transplanting, harvesting), and yield prediction
Authors: Aman Bhatta
The first layer of a deep CNN backbone applies filters to an image to extract the basic features available to later layers. During training, some filters may go inactive, mean ing all weights in the filter approach zero. An inactive fil ter in the final model represents a missed opportunity to extract a useful feature. This phenomenon is especially prevalent in specialized CNNs such as for face recogni tion (as opposed to, e.g., ImageNet). For example, in one the most widely face recognition model (ArcFace), about half of the convolution filters in the first layer are inactive. We propose a novel approach designed and tested specif ically for face recognition networks, known as "CRAFT: Contextual Re-Activation of Filters for Face Recognition Training". CRAFT identifies inactive filters during training and reinitializes them based on the context of strong filters at that stage in training. We show that CRAFT reduces fraction of inactive filters from 44% to 32% on average and discovers filter patterns not found by standard training. Compared to standard training without reactivation, CRAFT demonstrates enhanced model accuracy on standard face-recognition benchmark datasets including AgeDB-30, CPLFW, LFW, CALFW, and CFP-FP, as well as on more challenging datasets like IJBB and IJBC.
Authors: Shakiba Kheradmand, Daniel Rebain, Gopal Sharma, Hossam Isack, Abhishek Kar, Andrea Tagliasacchi, Kwang Moo Yi
We present an approach to accelerate Neural Field training by efficiently selecting sampling locations. While Neural Fields have recently become popular, it is often trained by uniformly sampling the training domain, or through handcrafted heuristics. We show that improved convergence and final training quality can be achieved by a soft mining technique based on importance sampling: rather than either considering or ignoring a pixel completely, we weigh the corresponding loss by a scalar. To implement our idea we use Langevin Monte-Carlo sampling. We show that by doing so, regions with higher error are being selected more frequently, leading to more than 2x improvement in convergence speed. The code and related resources for this study are publicly available at https://ubc-vision.github.io/nf-soft-mining/.
Authors: Zhonghao Wang, Wei Wei, Yang Zhao, Zhisheng Xiao, Mark Hasegawa-Johnson, Humphrey Shi, Tingbo Hou
This paper explores advancements in high-fidelity personalized image generation through the utilization of pre-trained text-to-image diffusion models. While previous approaches have made significant strides in generating versatile scenes based on text descriptions and a few input images, challenges persist in maintaining the subject fidelity within the generated images. In this work, we introduce an innovative algorithm named HiFi Tuner to enhance the appearance preservation of objects during personalized image generation. Our proposed method employs a parameter-efficient fine-tuning framework, comprising a denoising process and a pivotal inversion process. Key enhancements include the utilization of mask guidance, a novel parameter regularization technique, and the incorporation of step-wise subject representations to elevate the sample fidelity. Additionally, we propose a reference-guided generation approach that leverages the pivotal inversion of a reference image to mitigate unwanted subject variations and artifacts. We further extend our method to a novel image editing task: substituting the subject in an image through textual manipulations. Experimental evaluations conducted on the DreamBooth dataset using the Stable Diffusion model showcase promising results. Fine-tuning solely on textual embeddings improves CLIP-T score by 3.6 points and improves DINO score by 9.6 points over Textual Inversion. When fine-tuning all parameters, HiFi Tuner improves CLIP-T score by 1.2 points and improves DINO score by 1.2 points over DreamBooth, establishing a new state of the art.
Authors: Wujian Peng, Sicheng Xie, Zuyao You, Shiyi Lan, Zuxuan Wu
Vision language models (VLM) have demonstrated remarkable performance across various downstream tasks. However, understanding fine-grained visual-linguistic concepts, such as attributes and inter-object relationships, remains a significant challenge. While several benchmarks aim to evaluate VLMs in finer granularity, their primary focus remains on the linguistic aspect, neglecting the visual dimension. Here, we highlight the importance of evaluating VLMs from both a textual and visual perspective. We introduce a progressive pipeline to synthesize images that vary in a specific attribute while ensuring consistency in all other aspects. Utilizing this data engine, we carefully design a benchmark, SPEC, to diagnose the comprehension of object size, position, existence, and count. Subsequently, we conduct a thorough evaluation of four leading VLMs on SPEC. Surprisingly, their performance is close to random guess, revealing significant limitations. With this in mind, we propose a simply yet effective approach to optimize VLMs in fine-grained understanding, achieving significant improvements on SPEC without compromising the zero-shot performance. Results on two additional fine-grained benchmarks also show consistent improvements, further validating the transferability of our approach.
Authors: Ruoran Li, Runzhao Yang, Wenxin Xiang, Yuxiao Cheng, Tingxiong Xiao, Jinli Suo
Functional Magnetic Resonance Imaging (fMRI) data is a kind of widely used four-dimensional biomedical data, demanding effective compression but presenting unique challenges for compression due to its intricate temporal dynamics, low signal-to-noise ratio, and complicated underlying redundancies. This paper reports a novel compression paradigm specifically tailored for fMRI data based on Implicit Neural Representation (INR). The proposed approach focuses on removing the various redundancies among the time series, including (i) conducting spatial correlation modeling for intra-region dynamics, (ii) decomposing reusable neuronal activation patterns, and using proper initialization together with nonlinear fusion to describe the inter-region similarity. The above scheme properly incorporates the unique features of fMRI data, and experimental results on publicly available datasets demonstrate the effectiveness of the proposed method, surpassing state-of-the-art algorithms in both conventional image quality evaluation metrics and fMRI downstream tasks. This work in this paper paves the way for sharing massive fMRI data at low bandwidth and high fidelity.
Authors: Pilhyeon Lee, Hyeran Byun
Temporal sentence grounding aims to localize moments relevant to a language description. Recently, DETR-like approaches have shown notable progress by decoding the center and length of a target moment from learnable queries. However, they suffer from the issue of center misalignment raised by the inherent ambiguity of moment centers, leading to inaccurate predictions. To remedy this problem, we introduce a novel boundary-oriented moment formulation. In our paradigm, the model no longer needs to find the precise center but instead suffices to predict any anchor point within the interval, from which the onset and offset are directly estimated. Based on this idea, we design a Boundary-Aligned Moment Detection Transformer (BAM-DETR), equipped with a dual-pathway decoding process. Specifically, it refines the anchor and boundaries within parallel pathways using global and boundary-focused attention, respectively. This separate design allows the model to focus on desirable regions, enabling precise refinement of moment predictions. Further, we propose a quality-based ranking method, ensuring that proposals with high localization qualities are prioritized over incomplete ones. Extensive experiments verify the advantages of our methods, where our model records new state-of-the-art results on three benchmarks. Code is at https://github.com/Pilhyeon/BAM-DETR.
Authors: Zhengyue Zhao, Jinhao Duan, Kaidi Xu, Chenan Wang, Rui Zhangp Zidong Dup Qi Guo, Xing Hu
Stable Diffusion has established itself as a foundation model in generative AI artistic applications, receiving widespread research and application. Some recent fine-tuning methods have made it feasible for individuals to implant personalized concepts onto the basic Stable Diffusion model with minimal computational costs on small datasets. However, these innovations have also given rise to issues like facial privacy forgery and artistic copyright infringement. In recent studies, researchers have explored the addition of imperceptible adversarial perturbations to images to prevent potential unauthorized exploitation and infringements when personal data is used for fine-tuning Stable Diffusion. Although these studies have demonstrated the ability to protect images, it is essential to consider that these methods may not be entirely applicable in real-world scenarios. In this paper, we systematically evaluate the use of perturbations to protect images within a practical threat model. The results suggest that these approaches may not be sufficient to safeguard image privacy and copyright effectively. Furthermore, we introduce a purification method capable of removing protected perturbations while preserving the original image structure to the greatest extent possible. Experiments reveal that Stable Diffusion can effectively learn from purified images over all protective methods.
Authors: Yiwei Ma, Yijun Fan, Jiayi Ji, Haowei Wang, Xiaoshuai Sun, Guannan Jiang, Annan Shu, Rongrong Ji
In recent times, automatic text-to-3D content creation has made significant progress, driven by the development of pretrained 2D diffusion models. Existing text-to-3D methods typically optimize the 3D representation to ensure that the rendered image aligns well with the given text, as evaluated by the pretrained 2D diffusion model. Nevertheless, a substantial domain gap exists between 2D images and 3D assets, primarily attributed to variations in camera-related attributes and the exclusive presence of foreground objects. Consequently, employing 2D diffusion models directly for optimizing 3D representations may lead to suboptimal outcomes. To address this issue, we present X-Dreamer, a novel approach for high-quality text-to-3D content creation that effectively bridges the gap between text-to-2D and text-to-3D synthesis. The key components of X-Dreamer are two innovative designs: Camera-Guided Low-Rank Adaptation (CG-LoRA) and Attention-Mask Alignment (AMA) Loss. CG-LoRA dynamically incorporates camera information into the pretrained diffusion models by employing camera-dependent generation for trainable parameters. This integration enhances the alignment between the generated 3D assets and the camera's perspective. AMA loss guides the attention map of the pretrained diffusion model using the binary mask of the 3D object, prioritizing the creation of the foreground object. This module ensures that the model focuses on generating accurate and detailed foreground objects. Extensive evaluations demonstrate the effectiveness of our proposed method compared to existing text-to-3D approaches. Our project webpage: https://xmuxiaoma666.github.io/Projects/X-Dreamer .
Authors: Chong Wang, Yuanhong Chen, Fengbei Liu, Davis James McCarthy, Helen Frazer, Gustavo Carneiro
Prototypical-part interpretable methods, e.g., ProtoPNet, enhance interpretability by connecting classification predictions to class-specific training prototypes, thereby offering an intuitive insight into their decision-making. Current methods rely on a discriminative classifier trained with point-based learning techniques that provide specific values for prototypes. Such prototypes have relatively low representation power due to their sparsity and potential redundancy, with each prototype containing no variability measure. In this paper, we present a new generative learning of prototype distributions, named Mixture of Gaussian-distributed Prototypes (MGProto), which are represented by Gaussian mixture models (GMM). Such an approach enables the learning of more powerful prototype representations since each learned prototype will own a measure of variability, which naturally reduces the sparsity given the spread of the distribution around each prototype, and we also integrate a prototype diversity objective function into the GMM optimisation to reduce redundancy. Incidentally, the generative nature of MGProto offers a new and effective way for detecting out-of-distribution samples. To improve the compactness of MGProto, we further propose to prune Gaussian-distributed prototypes with a low prior. Experiments on CUB-200-2011, Stanford Cars, Stanford Dogs, and Oxford-IIIT Pets datasets show that MGProto achieves state-of-the-art classification and OoD detection performances with encouraging interpretability results.
Authors: Gege Gao, Weiyang Liu, Anpei Chen, Andreas Geiger, Bernhard Schölkopf
As pretrained text-to-image diffusion models become increasingly powerful, recent efforts have been made to distill knowledge from these text-to-image pretrained models for optimizing a text-guided 3D model. Most of the existing methods generate a holistic 3D model from a plain text input. This can be problematic when the text describes a complex scene with multiple objects, because the vectorized text embeddings are inherently unable to capture a complex description with multiple entities and relationships. Holistic 3D modeling of the entire scene further prevents accurate grounding of text entities and concepts. To address this limitation, we propose GraphDreamer, a novel framework to generate compositional 3D scenes from scene graphs, where objects are represented as nodes and their interactions as edges. By exploiting node and edge information in scene graphs, our method makes better use of the pretrained text-to-image diffusion model and is able to fully disentangle different objects without image-level supervision. To facilitate modeling of object-wise relationships, we use signed distance fields as representation and impose a constraint to avoid inter-penetration of objects. To avoid manual scene graph creation, we design a text prompt for ChatGPT to generate scene graphs based on text inputs. We conduct both qualitative and quantitative experiments to validate the effectiveness of GraphDreamer in generating high-fidelity compositional 3D scenes with disentangled object entities.
Authors: Zhenyu Zhou, Defang Chen, Can Wang, Chun Chen
Sampling from diffusion models can be treated as solving the corresponding ordinary differential equations (ODEs), with the aim of obtaining an accurate solution with as few number of function evaluations (NFE) as possible. Recently, various fast samplers utilizing higher-order ODE solvers have emerged and achieved better performance than the initial first-order one. However, these numerical methods inherently result in certain approximation errors, which significantly degrades sample quality with extremely small NFE (e.g., around 5). In contrast, based on the geometric observation that each sampling trajectory almost lies in a two-dimensional subspace embedded in the ambient space, we propose Approximate MEan-Direction Solver (AMED-Solver) that eliminates truncation errors by directly learning the mean direction for fast diffusion sampling. Besides, our method can be easily used as a plugin to further improve existing ODE-based samplers. Extensive experiments on image synthesis with the resolution ranging from 32 to 256 demonstrate the effectiveness of our method. With only 5 NFE, we achieve 7.14 FID on CIFAR-10, 13.75 FID on ImageNet 64$\times$64, and 12.79 FID on LSUN Bedroom. Our code is available at https://github.com/zhyzhouu/amed-solver.
Authors: Tongjia Chen, Hongshan Yu, Zhengeng Yang, Zechuan Li, Wei Sun, Chen Chen
Due to the resource-intensive nature of training vision-language models on expansive video data, a majority of studies have centered on adapting pre-trained image-language models to the video domain. Dominant pipelines propose to tackle the visual discrepancies with additional temporal learners while overlooking the substantial discrepancy for web-scaled descriptive narratives and concise action category names, leading to less distinct semantic space and potential performance limitations. In this work, we prioritize the refinement of text knowledge to facilitate generalizable video recognition. To address the limitations of the less distinct semantic space of category names, we prompt a large language model (LLM) to augment action class names into Spatio-Temporal Descriptors thus bridging the textual discrepancy and serving as a knowledge base for general recognition. Moreover, to assign the best descriptors with different video instances, we propose Optimal Descriptor Solver, forming the video recognition problem as solving the optimal matching flow across frame-level representations and descriptors. Comprehensive evaluations in zero-shot, few-shot, and fully supervised video recognition highlight the effectiveness of our approach. Our best model achieves a state-of-the-art zero-shot accuracy of 75.1% on Kinetics-600.
Authors: Chen Long, Wenxiao Zhang, Zhe Chen, Haiping Wang, Yuan Liu, Zhen Cao, Zhen Dong, Bisheng Yang
We propose SparseDC, a model for Depth Completion of Sparse and non-uniform depth inputs. Unlike previous methods focusing on completing fixed distributions on benchmark datasets (e.g., NYU with 500 points, KITTI with 64 lines), SparseDC is specifically designed to handle depth maps with poor quality in real usage. The key contributions of SparseDC are two-fold. First, we design a simple strategy, called SFFM, to improve the robustness under sparse input by explicitly filling the unstable depth features with stable image features. Second, we propose a two-branch feature embedder to predict both the precise local geometry of regions with available depth values and accurate structures in regions with no depth. The key of the embedder is an uncertainty-based fusion module called UFFM to balance the local and long-term information extracted by CNNs and ViTs. Extensive indoor and outdoor experiments demonstrate the robustness of our framework when facing sparse and non-uniform input depths. The pre-trained model and code are available at https://github.com/WHU-USI3DV/SparseDC.
Authors: Mahendran Narayanan
Movies wield significant influence in our lives, playing a pivotal role in the tourism industry of any country. The inclusion of picturesque landscapes, waterfalls, and mountains as backdrops in films serves to enhance the allure of specific scenarios. Recognizing the impact of movies on tourism, this paper introduces a method for identifying tourist destinations featured in films. We propose the development of a deep learning model capable of recognizing these locations during movie viewing. The model is trained on a dataset comprising major tourism destinations worldwide. Through this research, the goal is to enable viewers to identify the real-world locations depicted in movie scenes, offering a novel way to connect cinema with global travel experiences.
Authors: Bonifaz Stuhr
Unsupervised representation learning aims at finding methods that learn representations from data without annotation-based signals. Abstaining from annotations not only leads to economic benefits but may - and to some extent already does - result in advantages regarding the representation's structure, robustness, and generalizability to different tasks. In the long run, unsupervised methods are expected to surpass their supervised counterparts due to the reduction of human intervention and the inherently more general setup that does not bias the optimization towards an objective originating from specific annotation-based signals. While major advantages of unsupervised representation learning have been recently observed in natural language processing, supervised methods still dominate in vision domains for most tasks. In this dissertation, we contribute to the field of unsupervised (visual) representation learning from three perspectives: (i) Learning representations: We design unsupervised, backpropagation-free Convolutional Self-Organizing Neural Networks (CSNNs) that utilize self-organization- and Hebbian-based learning rules to learn convolutional kernels and masks to achieve deeper backpropagation-free models. (ii) Evaluating representations: We build upon the widely used (non-)linear evaluation protocol to define pretext- and target-objective-independent metrics for measuring and investigating the objective function mismatch between various unsupervised pretext tasks and target tasks. (iii) Transferring representations: We contribute CARLANE, the first 3-way sim-to-real domain adaptation benchmark for 2D lane detection, and a method based on prototypical self-supervised learning. Finally, we contribute a content-consistent unpaired image-to-image translation method that utilizes masks, global and local discriminators, and similarity sampling to mitigate content inconsistencies.
Authors: Saurabh Farkya, Aswin Raghavan, Avi Ziskind
Most real-world applications that employ deep neural networks (DNNs) quantize them to low precision to reduce the compute needs. We present a method to improve the robustness of quantized DNNs to white-box adversarial attacks. We first tackle the limitation of deterministic quantization to fixed ``bins'' by introducing a differentiable Stochastic Quantizer (SQ). We explore the hypothesis that different quantizations may collectively be more robust than each quantized DNN. We formulate a training objective to encourage different quantized DNNs to learn different representations of the input image. The training objective captures diversity and accuracy via mutual information between ensemble members. Through experimentation, we demonstrate substantial improvement in robustness against $L_\infty$ attacks even if the attacker is allowed to backpropagate through SQ (e.g., > 50\% accuracy to PGD(5/255) on CIFAR10 without adversarial training), compared to vanilla DNNs as well as existing ensembles of quantized DNNs. We extend the method to detect attacks and generate robustness profiles in the adversarial information plane (AIP), towards a unified analysis of different threat models by correlating the MI and accuracy.
Authors: Tao Lu, Mulin Yu, Linning Xu, Yuanbo Xiangli, Limin Wang, Dahua Lin, Bo Dai
Neural rendering methods have significantly advanced photo-realistic 3D scene rendering in various academic and industrial applications. The recent 3D Gaussian Splatting method has achieved the state-of-the-art rendering quality and speed combining the benefits of both primitive-based representations and volumetric representations. However, it often leads to heavily redundant Gaussians that try to fit every training view, neglecting the underlying scene geometry. Consequently, the resulting model becomes less robust to significant view changes, texture-less area and lighting effects. We introduce Scaffold-GS, which uses anchor points to distribute local 3D Gaussians, and predicts their attributes on-the-fly based on viewing direction and distance within the view frustum. Anchor growing and pruning strategies are developed based on the importance of neural Gaussians to reliably improve the scene coverage. We show that our method effectively reduces redundant Gaussians while delivering high-quality rendering. We also demonstrates an enhanced capability to accommodate scenes with varying levels-of-detail and view-dependent observations, without sacrificing the rendering speed.
Authors: Rémi Kazmierczak, Eloïse Berthier, Goran Frehse, Gianni Franchi
In this paper, we introduce an explainable algorithm designed from a multi-modal foundation model, that performs fast and explainable image classification. Drawing inspiration from CLIP-based Concept Bottleneck Models (CBMs), our method creates a latent space where each neuron is linked to a specific word. Observing that this latent space can be modeled with simple distributions, we use a Mixture of Gaussians (MoG) formalism to enhance the interpretability of this latent space. Then, we introduce CLIP-QDA, a classifier that only uses statistical values to infer labels from the concepts. In addition, this formalism allows for both local and global explanations. These explanations come from the inner design of our architecture, our work is part of a new family of greybox models, combining performances of opaque foundation models and the interpretability of transparent models. Our empirical findings show that in instances where the MoG assumption holds, CLIP-QDA achieves similar accuracy with state-of-the-art methods CBMs. Our explanations compete with existing XAI methods while being faster to compute.
Authors: Agelos Kratimenos, Jiahui Lei, Kostas Daniilidis
Accurately and efficiently modeling dynamic scenes and motions is considered so challenging a task due to temporal dynamics and motion complexity. To address these challenges, we propose DynMF, a compact and efficient representation that decomposes a dynamic scene into a few neural trajectories. We argue that the per-point motions of a dynamic scene can be decomposed into a small set of explicit or learned trajectories. Our carefully designed neural framework consisting of a tiny set of learned basis queried only in time allows for rendering speed similar to 3D Gaussian Splatting, surpassing 120 FPS, while at the same time, requiring only double the storage compared to static scenes. Our neural representation adequately constrains the inherently underconstrained motion field of a dynamic scene leading to effective and fast optimization. This is done by biding each point to motion coefficients that enforce the per-point sharing of basis trajectories. By carefully applying a sparsity loss to the motion coefficients, we are able to disentangle the motions that comprise the scene, independently control them, and generate novel motion combinations that have never been seen before. We can reach state-of-the-art render quality within just 5 minutes of training and in less than half an hour, we can synthesize novel views of dynamic scenes with superior photorealistic quality. Our representation is interpretable, efficient, and expressive enough to offer real-time view synthesis of complex dynamic scene motions, in monocular and multi-view scenarios.
Authors: Ziyun Wang, Friedhelm Hamann, Kenneth Chaney, Wen Jiang, Guillermo Gallego, Kostas Daniilidis
We present ContinuityCam, a novel approach to generate a continuous video from a single static RGB image, using an event camera. Conventional cameras struggle with high-speed motion capture due to bandwidth and dynamic range limitations. Event cameras are ideal sensors to solve this problem because they encode compressed change information at high temporal resolution. In this work, we propose a novel task called event-based continuous color video decompression, pairing single static color frames and events to reconstruct temporally continuous videos. Our approach combines continuous long-range motion modeling with a feature-plane-based synthesis neural integration model, enabling frame prediction at arbitrary times within the events. Our method does not rely on additional frames except for the initial image, increasing, thus, the robustness to sudden light changes, minimizing the prediction latency, and decreasing the bandwidth requirement. We introduce a novel single objective beamsplitter setup that acquires aligned images and events and a novel and challenging Event Extreme Decompression Dataset (E2D2) that tests the method in various lighting and motion profiles. We thoroughly evaluate our method through benchmarking reconstruction as well as various downstream tasks. Our approach significantly outperforms the event- and image- based baselines in the proposed task.
Authors: Ziyun Wang, Jinyuan Guo, Kostas Daniilidis
Event cameras are a novel type of biologically inspired vision sensor known for their high temporal resolution, high dynamic range, and low power consumption. Because of these properties, they are well-suited for processing fast motions that require rapid reactions. Although event cameras have recently shown competitive performance in unsupervised optical flow estimation, performance in detecting independently moving objects (IMOs) is lacking behind, although event-based methods would be suited for this task based on their low latency and HDR properties. Previous approaches to event-based IMO segmentation have been heavily dependent on labeled data. However, biological vision systems have developed the ability to avoid moving objects through daily tasks without being given explicit labels. In this work, we propose the first event framework that generates IMO pseudo-labels using geometric constraints. Due to its unsupervised nature, our method can handle an arbitrary number of not predetermined objects and is easily scalable to datasets where expensive IMO labels are not readily available. We evaluate our approach on the EVIMO dataset and show that it performs competitively with supervised methods, both quantitatively and qualitatively.
Authors: Matthew Gwilliam, Michael Cogswell, Meng Ye, Karan Sikka, Abhinav Shrivastava, Ajay Divakaran
Existing long video retrieval systems are trained and tested in the paragraph-to-video retrieval regime, where every long video is described by a single long paragraph. This neglects the richness and variety of possible valid descriptions of a video, which could be described in moment-by-moment detail, or in a single phrase summary, or anything in between. To provide a more thorough evaluation of the capabilities of long video retrieval systems, we propose a pipeline that leverages state-of-the-art large language models to carefully generate a diverse set of synthetic captions for long videos. We validate this pipeline's fidelity via rigorous human inspection. We then benchmark a representative set of video language models on these synthetic captions using a few long video datasets, showing that they struggle with the transformed data, especially the shortest captions. We also propose a lightweight fine-tuning method, where we use a contrastive loss to learn a hierarchical embedding loss based on the differing levels of information among the various captions. Our method improves performance both on the downstream paragraph-to-video retrieval task (+1.1% R@1 on ActivityNet), as well as for the various long video retrieval metrics we compute using our synthetic data (+3.6% R@1 for short descriptions on ActivityNet). For data access and other details, please refer to our project website at https://mgwillia.github.io/10k-words.
Authors: Or Greenberg, Eran Kishon, Dani Lischinski
Image-to-image translation (I2IT) refers to the process of transforming images from a source domain to a target domain while maintaining a fundamental connection in terms of image content. In the past few years, remarkable advancements in I2IT were achieved by Generative Adversarial Networks (GANs), which nevertheless struggle with translations requiring high precision. Recently, Diffusion Models have established themselves as the engine of choice for image generation. In this paper we introduce S2ST, a novel framework designed to accomplish global I2IT in complex photorealistic images, such as day-to-night or clear-to-rain translations of automotive scenes. S2ST operates within the seed space of a Latent Diffusion Model, thereby leveraging the powerful image priors learned by the latter. We show that S2ST surpasses state-of-the-art GAN-based I2IT methods, as well as diffusion-based approaches, for complex automotive scenes, improving fidelity while respecting the target domain's appearance across a variety of domains. Notably, S2ST obviates the necessity for training domain-specific translation networks.
Authors: Meera Hahn, Amit Raj, James M. Rehg
The challenging task of Vision-and-Language Navigation (VLN) requires embodied agents to follow natural language instructions to reach a goal location or object (e.g. `walk down the hallway and turn left at the piano'). For agents to complete this task successfully, they must be able to ground objects referenced into the instruction (e.g.`piano') into the visual scene as well as ground directional phrases (e.g.`turn left') into actions. In this work we ask the following question -- to what degree are spatial and directional language cues informing the navigation model's decisions? We propose a series of simple masking experiments to inspect the model's reliance on different parts of the instruction. Surprisingly we uncover that certain top performing models rely only on the noun tokens of the instructions. We propose two training methods to alleviate this concerning limitation.
Authors: Benjamin Schneider, Nils Lukas, Florian Kerschbaum
Web-scraped datasets are vulnerable to data poisoning, which can be used for backdooring deep image classifiers during training. Since training on large datasets is expensive, a model is trained once and re-used many times. Unlike adversarial examples, backdoor attacks often target specific classes rather than any class learned by the model. One might expect that targeting many classes through a naive composition of attacks vastly increases the number of poison samples. We show this is not necessarily true and more efficient, universal data poisoning attacks exist that allow controlling misclassifications from any source class into any target class with a small increase in poison samples. Our idea is to generate triggers with salient characteristics that the model can learn. The triggers we craft exploit a phenomenon we call inter-class poison transferability, where learning a trigger from one class makes the model more vulnerable to learning triggers for other classes. We demonstrate the effectiveness and robustness of our universal backdoor attacks by controlling models with up to 6,000 classes while poisoning only 0.15% of the training dataset.
Authors: Reza Kakavand, Mehrdad Palizi, Peyman Tahghighi, Reza Ahmadi, Neha Gianchandani, Samer Adeeb, Roberto Souza, W. Brent Edwards, Amin Komeili
Simulation studies like finite element (FE) modeling provide insight into knee joint mechanics without patient experimentation. Generic FE models represent biomechanical behavior of the tissue by overlooking variations in geometry, loading, and material properties of a population. On the other hand, subject-specific models include these specifics, resulting in enhanced predictive precision. However, creating such models is laborious and time-intensive. The present study aimed to enhance subject-specific knee joint FE modeling by incorporating a semi-automated segmentation algorithm. This segmentation was a 3D Swin UNETR for an initial segmentation of the femur and tibia, followed by a statistical shape model (SSM) adjustment to improve surface roughness and continuity. Five hundred and seven magnetic resonance images (MRIs) from the Osteoarthritis Initiative (OAI) database were used to build and validate the segmentation model. A semi-automated FE model was developed using this semi-automated segmentation. On the other hand, a manual FE model was developed through manual segmentation (i.e., the gold standard approach). Both FE models were subjected to gait loading. The predicted mechanical response of manual and semi-automated FE models were compared. In the result, our semi-automated segmentation achieved Dice similarity coefficient (DSC) over 98% for both femur and tibia. The mechanical results (max principal stress, max principal strain, fluid pressure, fibril strain, and contact area) showed no significant differences between the manual and semi-automated FE models, indicating the effectiveness of the proposed semi-automated segmentation in creating accurate knee joint FE models. ( https://data.mendeley.com/datasets/k5hdc9cz7w/1 ).
Authors: Bilel Tarchoun, Quazi Mishkatul Alam, Nael Abu-Ghazaleh, Ihsen Alouani
Adversarial patches exemplify the tangible manifestation of the threat posed by adversarial attacks on Machine Learning (ML) models in real-world scenarios. Robustness against these attacks is of the utmost importance when designing computer vision applications, especially for safety-critical domains such as CCTV systems. In most practical situations, monitoring open spaces requires multi-view systems to overcome acquisition challenges such as occlusion handling. Multiview object systems are able to combine data from multiple views, and reach reliable detection results even in difficult environments. Despite its importance in real-world vision applications, the vulnerability of multiview systems to adversarial patches is not sufficiently investigated. In this paper, we raise the following question: Does the increased performance and information sharing across views offer as a by-product robustness to adversarial patches? We first conduct a preliminary analysis showing promising robustness against off-the-shelf adversarial patches, even in an extreme setting where we consider patches applied to all views by all persons in Wildtrack benchmark. However, we challenged this observation by proposing two new attacks: (i) In the first attack, targeting a multiview CNN, we maximize the global loss by proposing gradient projection to the different views and aggregating the obtained local gradients. (ii) In the second attack, we focus on a Transformer-based multiview framework. In addition to the focal loss, we also maximize the transformer-specific loss by dissipating its attention blocks. Our results show a large degradation in the detection performance of victim multiview systems with our first patch attack reaching an attack success rate of 73% , while our second proposed attack reduced the performance of its target detector by 62%
Authors: Gokul Srinivasagan, Michael Deisher, Munir Georges
People with visual impairments have difficulty accessing touchscreen-enabled personal computing devices like mobile phones and laptops. The image-to-speech (ITS) systems can assist them in mitigating this problem, but their huge model size makes it extremely hard to be deployed on low-resourced embedded devices. In this paper, we aim to overcome this challenge by developing an efficient endto-end neural architecture for generating audio from tiny segments of display content on low-resource devices. We introduced a vision transformers-based image encoder and utilized knowledge distillation to compress the model from 6.1 million to 2.46 million parameters. Human and automatic evaluation results show that our approach leads to a very minimal drop in performance and can speed up the inference time by 22%.
Authors: Anusha Guruprasad
The classification of galaxies as spirals or ellipticals is a crucial task in understanding their formation and evolution. With the arrival of large-scale astronomical surveys, such as the Sloan Digital Sky Survey (SDSS), astronomers now have access to images of a vast number of galaxies. However, the visual inspection of these images is an impossible task for humans due to the sheer number of galaxies to be analyzed. To solve this problem, the Galaxy Zoo project was created to engage thousands of citizen scientists to classify the galaxies based on their visual features. In this paper, we present a machine learning model for galaxy classification using numerical data from the Galaxy Zoo[5] project. Our model utilizes a convolutional neural network architecture to extract features from galaxy images and classify them into spirals or ellipticals. We demonstrate the effectiveness of our model by comparing its performance with that of human classifiers using a subset of the Galaxy Zoo dataset. Our results show that our model achieves high accuracy in classifying galaxies and has the potential to significantly enhance our understanding of the formation and evolution of galaxies.
Authors: Naga VS Raviteja Chappa, Pha Nguyen, Page Daniel Dobbs, Khoa Luu
Group Activity Recognition (GAR) is a fundamental problem in computer vision, with diverse applications in sports video analysis, video surveillance, and social scene understanding. Unlike conventional action recognition, GAR aims to classify the actions of a group of individuals as a whole, requiring a deep understanding of their interactions and spatiotemporal relationships. To address the challenges in GAR, we present REACT (\textbf{R}ecognize \textbf{E}very \textbf{Act}ion Everywhere All At Once), a novel architecture inspired by the transformer encoder-decoder model explicitly designed to model complex contextual relationships within videos, including multi-modality and spatio-temporal features. Our architecture features a cutting-edge Vision-Language Encoder block for integrated temporal, spatial, and multi-modal interaction modeling. This component efficiently encodes spatiotemporal interactions, even with sparsely sampled frames, and recovers essential local information. Our Action Decoder Block refines the joint understanding of text and video data, allowing us to precisely retrieve bounding boxes, enhancing the link between semantics and visual reality. At the core, our Actor Fusion Block orchestrates a fusion of actor-specific data and textual features, striking a balance between specificity and context. Our method outperforms state-of-the-art GAR approaches in extensive experiments, demonstrating superior accuracy in recognizing and understanding group activities. Our architecture's potential extends to diverse real-world applications, offering empirical evidence of its performance gains. This work significantly advances the field of group activity recognition, providing a robust framework for nuanced scene comprehension.
Authors: Renos Zabounidis, Ini Oguntola, Konghao Zhao, Joseph Campbell, Simon Stepputtis, Katia Sycara
Concept bottleneck models (CBMs) are interpretable models that first predict a set of semantically meaningful features, i.e., concepts, from observations that are subsequently used to condition a downstream task. However, the model's performance strongly depends on the engineered features and can severely suffer from incomplete sets of concepts. Prior works have proposed a side channel -- a residual -- that allows for unconstrained information flow to the downstream task, thus improving model performance but simultaneously introducing information leakage, which is undesirable for interpretability. This work proposes three novel approaches to mitigate information leakage by disentangling concepts and residuals, investigating the critical balance between model performance and interpretability. Through extensive empirical analysis on the CUB, OAI, and CIFAR 100 datasets, we assess the performance of each disentanglement method and provide insights into when they work best. Further, we show how each method impacts the ability to intervene over the concepts and their subsequent impact on task performance.
Authors: Davide Cozzolino, Giovanni Poggi, Riccardo Corvi, Matthias Nießner, Luisa Verdoliva
Aim of this work is to explore the potential of pre-trained vision-language models (VLMs) for universal detection of AI-generated images. We develop a lightweight detection strategy based on CLIP features and study its performance in a wide variety of challenging scenarios. We find that, unlike previous belief, it is neither necessary nor convenient to use a large domain-specific dataset for training. On the contrary, by using only a handful of example images from a single generative model, a CLIP-based detector exhibits a surprising generalization ability and high robustness across several different architectures, including recent commercial tools such as Dalle-3, Midjourney v5, and Firefly. We match the SoTA on in-distribution data, and improve largely above it in terms of generalization to out-of-distribution data (+6% in terms of AUC) and robustness to impaired/laundered data (+13%). Our project is available at https://grip-unina.github.io/ClipBased-SyntheticImageDetection/
Authors: Kunyi Li, Michael Niemeyer, Nassir Navab, Federico Tombari
In recent years, coordinate-based neural implicit representations have shown promising results for the task of Simultaneous Localization and Mapping (SLAM). While achieving impressive performance on small synthetic scenes, these methods often suffer from oversmoothed reconstructions, especially for complex real-world scenes. In this work, we introduce DNS SLAM, a novel neural RGB-D semantic SLAM approach featuring a hybrid representation. Relying only on 2D semantic priors, we propose the first semantic neural SLAM method that trains class-wise scene representations while providing stable camera tracking at the same time. Our method integrates multi-view geometry constraints with image-based feature extraction to improve appearance details and to output color, density, and semantic class information, enabling many downstream applications. To further enable real-time tracking, we introduce a lightweight coarse scene representation which is trained in a self-supervised manner in latent space. Our experimental results achieve state-of-the-art performance on both synthetic data and real-world data tracking while maintaining a commendable operational speed on off-the-shelf hardware. Further, our method outputs class-wise decomposed reconstructions with better texture capturing appearance and geometric details.
Authors: Haolin Xiong, Sairisheek Muttukuru, Rishi Upadhyay, Pradyumna Chari, Achuta Kadambi
The problem of novel view synthesis has grown significantly in popularity recently with the introduction of Neural Radiance Fields (NeRFs) and other implicit scene representation methods. A recent advance, 3D Gaussian Splatting (3DGS), leverages an explicit representation to achieve real-time rendering with high-quality results. However, 3DGS still requires an abundance of training views to generate a coherent scene representation. In few shot settings, similar to NeRF, 3DGS tends to overfit to training views, causing background collapse and excessive floaters, especially as the number of training views are reduced. We propose a method to enable training coherent 3DGS-based radiance fields of 360 scenes from sparse training views. We find that using naive depth priors is not sufficient and integrate depth priors with generative and explicit constraints to reduce background collapse, remove floaters, and enhance consistency from unseen viewpoints. Experiments show that our method outperforms base 3DGS by up to 30.5% and NeRF-based methods by up to 15.6% in LPIPS on the MipNeRF-360 dataset with substantially less training and inference cost.
Authors: Jinxin Zhou, Tianyu Ding, Tianyi Chen, Jiachen Jiang, Ilya Zharkov, Zhihui Zhu, Luming Liang
We present DREAM, a novel training framework representing Diffusion Rectification and Estimation-Adaptive Models, requiring minimal code changes (just three lines) yet significantly enhancing the alignment of training with sampling in diffusion models. DREAM features two components: diffusion rectification, which adjusts training to reflect the sampling process, and estimation adaptation, which balances perception against distortion. When applied to image super-resolution (SR), DREAM adeptly navigates the tradeoff between minimizing distortion and preserving high image quality. Experiments demonstrate DREAM's superiority over standard diffusion-based SR methods, showing a $2$ to $3\times $ faster training convergence and a $10$ to $20\times$ reduction in necessary sampling steps to achieve comparable or superior results. We hope DREAM will inspire a rethinking of diffusion model training paradigms.
Authors: Linzi Xing, Quan Tran, Fabian Caba, Franck Dernoncourt, Seunghyun Yoon, Zhaowen Wang, Trung Bui, Giuseppe Carenini
Video topic segmentation unveils the coarse-grained semantic structure underlying videos and is essential for other video understanding tasks. Given the recent surge in multi-modal, relying solely on a single modality is arguably insufficient. On the other hand, prior solutions for similar tasks like video scene/shot segmentation cater to short videos with clear visual shifts but falter for long videos with subtle changes, such as livestreams. In this paper, we introduce a multi-modal video topic segmenter that utilizes both video transcripts and frames, bolstered by a cross-modal attention mechanism. Furthermore, we propose a dual-contrastive learning framework adhering to the unsupervised domain adaptation paradigm, enhancing our model's adaptability to longer, more semantically complex videos. Experiments on short and long video corpora demonstrate that our proposed solution, significantly surpasses baseline methods in terms of both accuracy and transferability, in both intra- and cross-domain settings.
Authors: Mena Shenouda, Eyjólfur Gudmundsson, Feng Li, Christopher M. Straus, Hedy L. Kindler, Arkadiusz Z. Dudek, Thomas Stinchcombe, Xiaofei Wang, Adam Starkey, Samuel G. Armato III
Malignant pleural mesothelioma (MPM) is the most common form of mesothelioma. To assess response to treatment, tumor measurements are acquired and evaluated based on a patient's longitudinal computed tomography (CT) scans. Tumor volume, however, is the more accurate metric for assessing tumor burden and response. Automated segmentation methods using deep learning can be employed to acquire volume, which otherwise is a tedious task performed manually. The deep learning-based tumor volume and contours can then be compared with a standard reference to assess the robustness of the automated segmentations. The purpose of this study was to evaluate the impact of probability map threshold on MPM tumor delineations generated using a convolutional neural network (CNN). Eighty-eight CT scans from 21 MPM patients were segmented by a VGG16/U-Net CNN. A radiologist modified the contours generated at a 0.5 probability threshold. Percent difference of tumor volume and overlap using the Dice Similarity Coefficient (DSC) were compared between the standard reference provided by the radiologist and CNN outputs for thresholds ranging from 0.001 to 0.9. CNN annotations consistently yielded smaller tumor volumes than radiologist contours. Reducing the probability threshold from 0.5 to 0.1 decreased the absolute percent volume difference, on average, from 43.96% to 24.18%. Median and mean DSC ranged from 0.58 to 0.60, with a peak at a threshold of 0.5; no distinct threshold was found for percent volume difference. No single output threshold in the CNN probability maps was optimal for both tumor volume and DSC. This work underscores the need to assess tumor volume and spatial overlap when evaluating CNN performance. While automated segmentations may yield comparable tumor volumes to that of the reference standard, the spatial region delineated by the CNN at a specific threshold is equally important.
Authors: Imane Koulali, M. Taner Eskil
In this study, we propose a novel motif-based approach for unsupervised textile anomaly detection that combines the benefits of traditional convolutional neural networks with those of an unsupervised learning paradigm. It consists of five main steps: preprocessing, automatic pattern period extraction, patch extraction, features selection and anomaly detection. This proposed approach uses a new dynamic and heuristic method for feature selection which avoids the drawbacks of initialization of the number of filters (neurons) and their weights, and those of the backpropagation mechanism such as the vanishing gradients, which are common practice in the state-of-the-art methods. The design and training of the network are performed in a dynamic and input domain-based manner and, thus, no ad-hoc configurations are required. Before building the model, only the number of layers and the stride are defined. We do not initialize the weights randomly nor do we define the filter size or number of filters as conventionally done in CNN-based approaches. This reduces effort and time spent on hyperparameter initialization and fine-tuning. Only one defect-free sample is required for training and no further labeled data is needed. The trained network is then used to detect anomalies on defective fabric samples. We demonstrate the effectiveness of our approach on the Patterned Fabrics benchmark dataset. Our algorithm yields reliable and competitive results (on recall, precision, accuracy and f1- measure) compared to state-of-the-art unsupervised approaches, in less time, with efficient training in a single epoch and a lower computational cost.
Authors: Xuan-Bac Nguyen, Xin Li, Samee U. Khan, Khoa Luu
"Perception is reality". Human perception plays a vital role in forming beliefs and understanding reality. Exploring how the human brain works in the visual system facilitates bridging the gap between human visual perception and computer vision models. However, neuroscientists study the brain via Neuroimaging, i.e., Functional Magnetic Resonance Imaging (fMRI), to discover the brain's functions. These approaches face interpretation challenges where fMRI data can be complex and require expertise. Therefore, neuroscientists make inferences about cognitive processes based on patterns of brain activities, which can lead to potential misinterpretation or limited functional understanding. In this work, we first present a simple yet effective Brainformer approach, a novel Transformer-based framework, to analyze the patterns of fMRI in the human perception system from the machine learning perspective. Secondly, we introduce a novel mechanism incorporating fMRI, which represents the human brain activities, as the supervision for the machine vision model. This work also introduces a novel perspective on transferring knowledge from human perception to neural networks. Through our experiments, we demonstrated that by leveraging fMRI information, the machine vision model can achieve potential results compared to the current State-of-the-art methods in various image recognition tasks.
Authors: Zhuoran Zheng, Boxue Xiao
Currently, to further improve visual enjoyment, Ultra-High-Definition (UHD) images are catching wide attention. Here, UHD images are usually referred to as having a resolution greater than or equal to $3840 \times 2160$. However, since the imaging equipment is subject to environmental noise or equipment jitter, UHD images are prone to contrast degradation, blurring, low dynamic range, etc. To address these issues, a large number of algorithms for UHD image enhancement have been proposed. In this paper, we introduce the current state of UHD image enhancement from two perspectives, one is the application field and the other is the technology. In addition, we briefly explore its trends.
Authors: Haithem Turki, Michael Zollhöfer, Christian Richardt, Deva Ramanan
Neural Radiance Fields (NeRFs) can be dramatically accelerated by spatial grid representations. However, they do not explicitly reason about scale and so introduce aliasing artifacts when reconstructing scenes captured at different camera distances. Mip-NeRF and its extensions propose scale-aware renderers that project volumetric frustums rather than point samples but such approaches rely on positional encodings that are not readily compatible with grid methods. We propose a simple modification to grid-based models by training model heads at different spatial grid resolutions. At render time, we simply use coarser grids to render samples that cover larger volumes. Our method can be easily applied to existing accelerated NeRF methods and significantly improves rendering quality (reducing error rates by 20-90% across synthetic and unbounded real-world scenes) while incurring minimal performance overhead (as each model head is quick to evaluate). Compared to Mip-NeRF, we reduce error rates by 20% while training over 60x faster.
Authors: Hayden Moore
Computer Vision (CV) systems are increasingly being adopted into Command and Control (C2) systems to improve intelligence analysis on the battlefield, the tactical edge. CV systems leverage Artificial Intelligence (AI) algorithms to help visualize and interpret the environment, enhancing situational awareness. However, the adaptability of CV systems at the tactical edge remains challenging due to rapidly changing environments and objects which can confuse the deployed models. A CV model leveraged in this environment can become uncertain in its predictions, as the environment and the objects existing in the environment begin to change. Additionally, mission objectives can rapidly change leading to adjustments in technology, camera angles, and image resolutions. All of which can negatively affect the performance of and potentially introduce uncertainty into the system. When the training environment and/or technology differs from the deployment environment, CV models can perform unexpectedly. Unfortunately, most scenarios at the tactical edge do not incorporate Uncertainty Quantification (UQ) into their deployed C2 and CV systems. This concept paper explores the idea of synchronizing robust data operations and model fine-tuning driven by UQ all at the tactical edge. Specifically, curating datasets and training child models based on the residuals of predictions, using these child models to calculate prediction intervals (PI), and then using these PI to calibrate the deployed models. By incorporating UQ into the core operations surrounding C2 and CV systems at the tactical edge, we can help drive purposeful adaptability on the battlefield.
Authors: Longfeng Nie, Yuntian Chen, Dongxiao Zhang, Xinyue Liu, Wentian Yuan
Global horizontal irradiance (GHI) plays a vital role in estimating solar energy resources, which are used to generate sustainable green energy. In order to estimate GHI with high spatial resolution, a quantitative irradiance estimation network, named QIENet, is proposed. Specifically, the temporal and spatial characteristics of remote sensing data of the satellite Himawari-8 are extracted and fused by recurrent neural network (RNN) and convolution operation, respectively. Not only remote sensing data, but also GHI-related time information (hour, day, and month) and geographical information (altitude, longitude, and latitude), are used as the inputs of QIENet. The satellite spectral channels B07 and B11 - B15 and time are recommended as model inputs for QIENet according to the spatial distributions of annual solar energy. Meanwhile, QIENet is able to capture the impact of various clouds on hourly GHI estimates. More importantly, QIENet does not overestimate ground observations and can also reduce RMSE by 27.51%/18.00%, increase R2 by 20.17%/9.42%, and increase r by 8.69%/3.54% compared with ERA5/NSRDB. Furthermore, QIENet is capable of providing a high-fidelity hourly GHI database with spatial resolution 0.02{\deg} * 0.02{\deg}(approximately 2km * 2km) for many applied energy fields.
Authors: Niranjan Rajesh, Debayan Gupta
In the backdrop of increasing data requirements of Deep Neural Networks for object recognition that is growing more untenable by the day, we present Developmental PreTraining (DPT) as a possible solution. DPT is designed as a curriculum-based pre-training approach designed to rival traditional pre-training techniques that are data-hungry. These training approaches also introduce unnecessary features that could be misleading when the network is employed in a downstream classification task where the data is sufficiently different from the pre-training data and is scarce. We design the curriculum for DPT by drawing inspiration from human infant visual development. DPT employs a phased approach where carefully-selected primitive and universal features like edges and shapes are taught to the network participating in our pre-training regime. A model that underwent the DPT regime is tested against models with randomised weights to evaluate the viability of DPT.
Authors: Nikhel Gupta, Zeeshan Hayder, Ray P. Norris, Minh Huynh, Lars Petersson
Creating radio galaxy catalogues from next-generation deep surveys requires automated identification of associated components of extended sources and their corresponding infrared hosts. In this paper, we introduce RadioGalaxyNET, a multimodal dataset, and a suite of novel computer vision algorithms designed to automate the detection and localization 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 channels. Each instance provides information about the extended radio galaxy class, its corresponding bounding box encompassing all components, the pixel-level segmentation mask, and the keypoint position of its corresponding infrared host galaxy. RadioGalaxyNET is the first dataset to include images from the highly sensitive Australian Square Kilometre Array Pathfinder (ASKAP) radio telescope, corresponding infrared images, and instance-level annotations for galaxy detection. We benchmark several object detection algorithms on the dataset and propose a novel multimodal approach to simultaneously detect radio galaxies and the positions of infrared hosts.
Authors: Longfeng Nie, Yuntian Chen, Mengge Du, Changqi Sun, Dongxiao Zhang
Cloud types, as a type of meteorological data, are of particular significance for evaluating changes in rainfall, heatwaves, water resources, floods and droughts, food security and vegetation cover, as well as land use. In order to effectively utilize high-resolution geostationary observations, a knowledge-based data-driven (KBDD) framework for all-day identification of cloud types based on spectral information from Himawari-8/9 satellite sensors is designed. And a novel, simple and efficient network, named CldNet, is proposed. Compared with widely used semantic segmentation networks, including SegNet, PSPNet, DeepLabV3+, UNet, and ResUnet, our proposed model CldNet with an accuracy of 80.89+-2.18% is state-of-the-art in identifying cloud types and has increased by 32%, 46%, 22%, 2%, and 39%, respectively. With the assistance of auxiliary information (e.g., satellite zenith/azimuth angle, solar zenith/azimuth angle), the accuracy of CldNet-W using visible and near-infrared bands and CldNet-O not using visible and near-infrared bands on the test dataset is 82.23+-2.14% and 73.21+-2.02%, respectively. Meanwhile, the total parameters of CldNet are only 0.46M, making it easy for edge deployment. More importantly, the trained CldNet without any fine-tuning can predict cloud types with higher spatial resolution using satellite spectral data with spatial resolution 0.02{\deg}*0.02{\deg}, which indicates that CldNet possesses a strong generalization ability. In aggregate, the KBDD framework using CldNet is a highly effective cloud-type identification system capable of providing a high-fidelity, all-day, spatiotemporal cloud-type database for many climate assessment fields.
Authors: Zidu Wang, Xiangyu Zhu, Tianshuo Zhang, Baiqin Wang, Zhen Lei
3D Morphable Models (3DMMs) provide promising 3D face reconstructions in various applications. However, existing methods struggle to reconstruct faces with extreme expressions due to deficiencies in supervisory signals, such as sparse or inaccurate landmarks. Segmentation information contains effective geometric contexts for face reconstruction. Certain attempts intuitively depend on differentiable renderers to compare the rendered silhouettes of reconstruction with segmentation, which is prone to issues like local optima and gradient instability. In this paper, we fully utilize the facial part segmentation geometry by introducing Part Re-projection Distance Loss (PRDL). Specifically, PRDL transforms facial part segmentation into 2D points and re-projects the reconstruction onto the image plane. Subsequently, by introducing grid anchors and computing different statistical distances from these anchors to the point sets, PRDL establishes geometry descriptors to optimize the distribution of the point sets for face reconstruction. PRDL exhibits a clear gradient compared to the renderer-based methods and presents state-of-the-art reconstruction performance in extensive quantitative and qualitative experiments. The project will be publicly available.
Authors: Yiming Zhao, Tao Zhou, Yunqi Gu, Yi Zhou, Yizhe Zhang, Ye Wu, Huazhu Fu
Polyp segmentation plays a vital role in accurately locating polyps at an early stage, which holds significant clinical importance for the prevention of colorectal cancer. Various polyp segmentation methods have been developed using fully-supervised deep learning techniques. However, pixel-wise annotation for polyp images by physicians during the diagnosis is both time-consuming and expensive. Moreover, visual foundation models such as the Segment Anything Model (SAM) have shown remarkable performance. Nevertheless, directly applying SAM to medical segmentation may not produce satisfactory results due to the inherent absence of medical knowledge. In this paper, we propose a novel SAM-guided Collaborative Learning Network (SAM-CLNet) for scribble-supervised polyp segmentation, enabling a collaborative learning process between our segmentation network and SAM to boost the model performance. Specifically, we first propose a Cross-level Enhancement and Aggregation Network (CEA-Net) for weakly-supervised polyp segmentation. Within CEA-Net, we propose a Cross-level Enhancement Module (CEM) that integrates the adjacent features to enhance the representation capabilities of different resolution features. Additionally, a Feature Aggregation Module (FAM) is employed to capture richer features across multiple levels. Moreover, we present a box-augmentation strategy that combines the segmentation maps generated by CEA-Net with scribble annotations to create more precise prompts. These prompts are then fed into SAM, generating segmentation SAM-guided masks, which can provide additional supervision to train CEA-Net effectively. Furthermore, we present an Image-level Filtering Mechanism to filter out unreliable SAM-guided masks. Extensive experimental results show that our SAM-CLNet outperforms state-of-the-art weakly-supervised segmentation methods.
Authors: Seyedalireza Khoshsirat, Chandra Kambhamettu
Stein's paradox holds considerable sway in high-dimensional statistics, highlighting that the sample mean, traditionally considered the de facto estimator, might not be the most efficacious in higher dimensions. To address this, the James-Stein estimator proposes an enhancement by steering the sample means toward a more centralized mean vector. In this paper, first, we establish that normalization layers in deep learning use inadmissible estimators for mean and variance. Next, we introduce a novel method to employ the James-Stein estimator to improve the estimation of mean and variance within normalization layers. We evaluate our method on different computer vision tasks: image classification, semantic segmentation, and 3D object classification. Through these evaluations, it is evident that our improved normalization layers consistently yield superior accuracy across all tasks without extra computational burden. Moreover, recognizing that a plethora of shrinkage estimators surpass the traditional estimator in performance, we study two other prominent shrinkage estimators: Ridge and LASSO. Additionally, we provide visual representations to intuitively demonstrate the impact of shrinkage on the estimated layer statistics. Finally, we study the effect of regularization and batch size on our modified batch normalization. The studies show that our method is less sensitive to batch size and regularization, improving accuracy under various setups.
Authors: Dengbo Li, Jieren Cheng, Boyi Liu
In this work, we present a novel framework for camera relocation in autonomous vehicles, leveraging deep neural networks (DNN). While existing literature offers various DNN-based camera relocation methods, their deployment is hindered by their high computational demands during inference. In contrast, our approach addresses this challenge through edge cloud collaboration. Specifically, we strategically offload certain modules of the neural network to the server and evaluate the inference time of data frames under different network segmentation schemes to guide our offloading decisions. Our findings highlight the vital role of server-side offloading in DNN-based camera relocation for autonomous vehicles, and we also discuss the results of data fusion. Finally, we validate the effectiveness of our proposed framework through experimental evaluation.
Authors: Gongye Liu, Menghan Xia, Yong Zhang, Haoxin Chen, Jinbo Xing, Xintao Wang, Yujiu Yang, Ying Shan
Text-to-video (T2V) models have shown remarkable capabilities in generating diverse videos. However, they struggle to produce user-desired stylized videos due to (i) text's inherent clumsiness in expressing specific styles and (ii) the generally degraded style fidelity. To address these challenges, we introduce StyleCrafter, a generic method that enhances pre-trained T2V models with a style control adapter, enabling video generation in any style by providing a reference image. Considering the scarcity of stylized video datasets, we propose to first train a style control adapter using style-rich image datasets, then transfer the learned stylization ability to video generation through a tailor-made finetuning paradigm. To promote content-style disentanglement, we remove style descriptions from the text prompt and extract style information solely from the reference image using a decoupling learning strategy. Additionally, we design a scale-adaptive fusion module to balance the influences of text-based content features and image-based style features, which helps generalization across various text and style combinations. StyleCrafter efficiently generates high-quality stylized videos that align with the content of the texts and resemble the style of the reference images. Experiments demonstrate that our approach is more flexible and efficient than existing competitors.
Authors: Ziyu Zhou, Haozhe Luo, Jiaxuan Pang, Xiaowei Ding, Michael Gotway, Jianming Liang
Self-supervised learning (SSL) approaches have recently shown substantial success in learning visual representations from unannotated images. Compared with photographic images, medical images acquired with the same imaging protocol exhibit high consistency in anatomy. To exploit this anatomical consistency, this paper introduces a novel SSL approach, called PEAC (patch embedding of anatomical consistency), for medical image analysis. Specifically, in this paper, we propose to learn global and local consistencies via stable grid-based matching, transfer pre-trained PEAC models to diverse downstream tasks, and extensively demonstrate that (1) PEAC achieves significantly better performance than the existing state-of-the-art fully/self-supervised methods, and (2) PEAC captures the anatomical structure consistency across views of the same patient and across patients of different genders, weights, and healthy statuses, which enhances the interpretability of our method for medical image analysis.
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: Xiao Wang, Yaoyu Li, Tian Gan, Zheng Zhang, Jingjing Lv, Liqiang Nie
Recent advancements in video-language understanding have been established on the foundation of image-text models, resulting in promising outcomes due to the shared knowledge between images and videos. However, video-language understanding presents unique challenges due to the inclusion of highly complex semantic details, which result in information redundancy, temporal dependency, and scene complexity. Current techniques have only partially tackled these issues, and our quantitative analysis indicates that some of these methods are complementary. In light of this, we propose a novel framework called RTQ (Refine, Temporal model, and Query), which addresses these challenges simultaneously. The approach involves refining redundant information within frames, modeling temporal relations among frames, and querying task-specific information from the videos. Remarkably, our model demonstrates outstanding performance even in the absence of video-language pre-training, and the results are comparable with or superior to those achieved by state-of-the-art pre-training methods.
Authors: Anagha Deshpande, Vedant Deshpande
The recent advances in artificial intelligence and deep learning facilitate automation in various applications including home automation, smart surveillance systems, and healthcare among others. Human Activity Recognition is one of its emerging applications, which can be implemented in a classroom environment to enhance safety, efficiency, and overall educational quality. This paper proposes a system for detecting and recognizing the activities of students in a classroom environment. The dataset has been structured and recorded by the authors since a standard dataset for this task was not available at the time of this study. Transfer learning, a widely adopted method within the field of deep learning, has proven to be helpful in complex tasks like image and video processing. Pretrained models including VGG-16, ResNet-50, InceptionV3, and Xception are used for feature extraction and classification tasks. Xception achieved an accuracy of 93%, on the novel classroom dataset, outperforming the other three models in consideration. The system proposed in this study aims to introduce a safer and more productive learning environment for students and educators.
Authors: Haokun Chen, Xu Yang, Yuhang Huang, Zihan Wu, Jing Wang, Xin Geng
After pre-training by generating the next word conditional on previous words, the Language Model (LM) acquires the ability of In-Context Learning (ICL) that can learn a new task conditional on the context of the given in-context examples (ICEs). Similarly, visually-conditioned Language Modelling is also used to train Vision-Language Models (VLMs) with ICL ability. However, such VLMs typically exhibit weaker classification abilities compared to contrastive learning-based models like CLIP, since the Language Modelling objective does not directly contrast whether an object is paired with a text. To improve the ICL of classification, using more ICEs to provide more knowledge is a straightforward way. However, this may largely increase the selection time, and more importantly, the inclusion of additional in-context images tends to extend the length of the in-context sequence beyond the processing capacity of a VLM. To alleviate these limitations, we propose to manipulate the label space of each ICE to increase its knowledge density, allowing for fewer ICEs to convey as much information as a larger set would. Specifically, we propose two strategies which are Label Distribution Enhancement and Visual Descriptions Enhancement to improve In-context classification performance on diverse datasets, including the classic ImageNet and more fine-grained datasets like CUB-200. Specifically, using our approach on ImageNet, we increase accuracy from 74.70\% in a 4-shot setting to 76.21\% with just 2 shots. surpassing CLIP by 0.67\%. On CUB-200, our method raises 1-shot accuracy from 48.86\% to 69.05\%, 12.15\% higher than CLIP. The code is given in https://anonymous.4open.science/r/MLS_ICC.
Authors: Rohan Shad, Cyril Zakka, Dhamanpreet Kaur, Robyn Fong, Ross Warren Filice, John Mongan, Kimberly Kalianos, Nishith Khandwala, David Eng, Matthew Leipzig, Walter Witschey, Alejandro de Feria, Victor Ferrari, Euan Ashley, Michael A. Acker, Curtis Langlotz, William Hiesinger
Cardiac MRI allows for a comprehensive assessment of myocardial structure, function, and tissue characteristics. Here we describe a foundational vision system for cardiac MRI, capable of representing the breadth of human cardiovascular disease and health. Our deep learning model is trained via self-supervised contrastive learning, by which visual concepts in cine-sequence cardiac MRI scans are learned from the raw text of the accompanying radiology reports. We train and evaluate our model on data from four large academic clinical institutions in the United States. We additionally showcase the performance of our models on the UK BioBank, and two additional publicly available external datasets. We explore emergent zero-shot capabilities of our system, and demonstrate remarkable performance across a range of tasks; including the problem of left ventricular ejection fraction regression, and the diagnosis of 35 different conditions such as cardiac amyloidosis and hypertrophic cardiomyopathy. We show that our deep learning system is capable of not only understanding the staggering complexity of human cardiovascular disease, but can be directed towards clinical problems of interest yielding impressive, clinical grade diagnostic accuracy with a fraction of the training data typically required for such tasks.
Authors: Leting Zhouli, Peiyong Wang, Udaya Parampalli
In recent years, Classical Convolutional Neural Networks (CNNs) have been applied for image recognition successfully. Quantum Convolutional Neural Networks (QCNNs) are proposed as a novel generalization to CNNs by using quantum mechanisms. The quantum mechanisms lead to an efficient training process in QCNNs by reducing the size of input from $N$ to $log_2N$. This paper implements and compares both CNNs and QCNNs by testing losses and prediction accuracy on three commonly used datasets. The datasets include the MNIST hand-written digits, Fashion MNIST and cat/dog face images. Additionally, data augmentation (DA), a technique commonly used in CNNs to improve the performance of classification by generating similar images based on original inputs, is also implemented in QCNNs. Surprisingly, the results showed that data augmentation didn't improve QCNNs performance. The reasons and logic behind this result are discussed, hoping to expand our understanding of Quantum machine learning theory.
Authors: Shaohua Dong, Yunhe Feng, Qing Yang, Yan Huang, Dongfang Liu, Heng Fan
Multimodal (e.g., RGB-Depth/RGB-Thermal) fusion has shown great potential for improving semantic segmentation in complex scenes (e.g., indoor/low-light conditions). Existing approaches often fully fine-tune a dual-branch encoder-decoder framework with a complicated feature fusion strategy for achieving multimodal semantic segmentation, which is training-costly due to the massive parameter updates in feature extraction and fusion. To address this issue, we propose a surprisingly simple yet effective dual-prompt learning network (dubbed DPLNet) for training-efficient multimodal (e.g., RGB-D/T) semantic segmentation. The core of DPLNet is to directly adapt a frozen pre-trained RGB model to multimodal semantic segmentation, reducing parameter updates. For this purpose, we present two prompt learning modules, comprising multimodal prompt generator (MPG) and multimodal feature adapter (MFA). MPG works to fuse the features from different modalities in a compact manner and is inserted from shadow to deep stages to generate the multi-level multimodal prompts that are injected into the frozen backbone, while MPG adapts prompted multimodal features in the frozen backbone for better multimodal semantic segmentation. Since both the MPG and MFA are lightweight, only a few trainable parameters (3.88M, 4.4% of the pre-trained backbone parameters) are introduced for multimodal feature fusion and learning. Using a simple decoder (3.27M parameters), DPLNet achieves new state-of-the-art performance or is on a par with other complex approaches on four RGB-D/T semantic segmentation datasets while satisfying parameter efficiency. Moreover, we show that DPLNet is general and applicable to other multimodal tasks such as salient object detection and video semantic segmentation. Without special design, DPLNet outperforms many complicated models. Our code will be available at github.com/ShaohuaDong2021/DPLNet.
Authors: Ziyu Wang, Yue Xu, Cewu Lu, Yong-Lu Li
Recently, dataset distillation has paved the way towards efficient machine learning, especially for image datasets. However, the distillation for videos, characterized by an exclusive temporal dimension, remains an underexplored domain. In this work, we provide the first systematic study of video distillation and introduce a taxonomy to categorize temporal compression. Our investigation reveals that the temporal information is usually not well learned during distillation , and the temporal dimension of synthetic data contributes little. The observations motivate our unified framework of disentangling the dynamic and static information in the videos. It first distills the videos into still images as static memory and then compensates the dynamic and motion information with a learnable dynamic memory block. Our method achieves state-of-the-art on video datasets at different scales, with notably smaller storage expenditure. Our code will be publicly available.
Authors: Jiayi Li, Rohan Taori, Tatsunori B. Hashimoto
Active learning aims to enhance model performance by strategically labeling informative data points. While extensively studied, its effectiveness on large-scale, real-world datasets remains underexplored. Existing research primarily focuses on single-source data, ignoring the multi-domain nature of real-world data. We introduce a multi-domain active learning benchmark to bridge this gap. Our benchmark demonstrates that traditional single-domain active learning strategies are often less effective than random selection in multi-domain scenarios. We also introduce CLIP-GeoYFCC, a novel large-scale image dataset built around geographical domains, in contrast to existing genre-based domain datasets. Analysis on our benchmark shows that all multi-domain strategies exhibit significant tradeoffs, with no strategy outperforming across all datasets or all metrics, emphasizing the need for future research.
Authors: Yunjie Wu, Yapeng Meng, Zhipeng Hu, Lincheng Li, Haoqian Wu, Kun Zhou, Weiwei Xu, Xin Yu
Text-guided 3D face synthesis has achieved remarkable results by leveraging text-to-image (T2I) diffusion models. However, most existing works focus solely on the direct generation, ignoring the editing, restricting them from synthesizing customized 3D faces through iterative adjustments. In this paper, we propose a unified text-guided framework from face generation to editing. In the generation stage, we propose a geometry-texture decoupled generation to mitigate the loss of geometric details caused by coupling. Besides, decoupling enables us to utilize the generated geometry as a condition for texture generation, yielding highly geometry-texture aligned results. We further employ a fine-tuned texture diffusion model to enhance texture quality in both RGB and YUV space. In the editing stage, we first employ a pre-trained diffusion model to update facial geometry or texture based on the texts. To enable sequential editing, we introduce a UV domain consistency preservation regularization, preventing unintentional changes to irrelevant facial attributes. Besides, we propose a self-guided consistency weight strategy to improve editing efficacy while preserving consistency. Through comprehensive experiments, we showcase our method's superiority in face synthesis. Project page: https://faceg2e.github.io/.
Authors: Fangxin Shang, Jie Fu, Yehui Yang, Lei Ma
In the field of medical imaging, the scarcity of large-scale datasets due to privacy restrictions stands as a significant barrier to develop large models for medical. To address this issue, we introduce SynFundus-1M, a high-quality synthetic dataset with over 1 million retinal fundus images and extensive disease and pathologies annotations, which is generated by a Denoising Diffusion Probabilistic Model. The SynFundus-Generator and SynFundus-1M achieve superior Frechet Inception Distance (FID) scores compared to existing methods on main-stream public real datasets. Furthermore, the ophthalmologists evaluation validate the difficulty in discerning these synthetic images from real ones, confirming the SynFundus-1M's authenticity. Through extensive experiments, we demonstrate that both CNN and ViT can benifit from SynFundus-1M by pretraining or training directly. Compared to datasets like ImageNet or EyePACS, models train on SynFundus-1M not only achieve better performance but also faster convergence on various downstream tasks.
Authors: Maneesh John, Jyothi Rikhab Chand, Mathews Jacob
The recovery of magnetic resonance (MR) images from undersampled measurements is a key problem that has seen extensive research in recent years. Unrolled approaches, which rely on end-to-end training of convolutional neural network (CNN) blocks within iterative reconstruction algorithms, offer state-of-the-art performance. These algorithms require a large amount of memory during training, making them difficult to employ in high-dimensional applications. Deep equilibrium (DEQ) models and the recent monotone operator learning (MOL) approach were introduced to eliminate the need for unrolling, thus reducing the memory demand during training. Both approaches require a Lipschitz constraint on the network to ensure that the forward and backpropagation iterations converge. Unfortunately, the constraint often results in reduced performance compared to unrolled methods. The main focus of this work is to relax the constraint on the CNN block in two different ways. Inspired by convex-non-convex regularization strategies, we now impose the monotone constraint on the sum of the gradient of the data term and the CNN block, rather than constrain the CNN itself to be a monotone operator. This approach enables the CNN to learn possibly non-monotone score functions, which can translate to improved performance. In addition, we only restrict the operator to be monotone in a local neighborhood around the image manifold. Our theoretical results show that the proposed algorithm is guaranteed to converge to the fixed point and that the solution is robust to input perturbations, provided that it is initialized close to the true solution. Our empirical results show that the relaxed constraints translate to improved performance and that the approach enjoys robustness to input perturbations similar to MOL.
Authors: Yuxia Huang, Zhonghui Wu, Xiaoling Xu, Minghui Zhang, Shanshan Wang, Qiegen Liu
Multi-contrast magnetic resonance imaging is a significant and essential medical imaging technique.However, multi-contrast imaging has longer acquisition time and is easy to cause motion artifacts. In particular, the acquisition time for a T2-weighted image is prolonged due to its longer repetition time (TR). On the contrary, T1-weighted image has a shorter TR. Therefore,utilizing complementary information across T1 and T2-weighted image is a way to decrease the overall imaging time. Previous T1-assisted T2 reconstruction methods have mostly focused on image domain using whole-based image fusion approaches. The image domain reconstruction method has the defects of high computational complexity and limited flexibility. To address this issue, we propose a novel multi-contrast imaging method called partition-based k-space synthesis (PKS) which can achieve super reconstruction quality of T2-weighted image by feature fusion. Concretely, we first decompose fully-sampled T1 k-space data and under-sampled T2 k-space data into two sub-data, separately. Then two new objects are constructed by combining the two sub-T1/T2 data. After that, the two new objects as the whole data to realize the reconstruction of T2-weighted image. Finally, the objective T2 is synthesized by extracting the sub-T2 data of each part. Experimental results showed that our combined technique can achieve comparable or better results than using traditional k-space parallel imaging(SAKE) that processes each contrast independently.
Authors: Kshitij Deshpande, Varad Mashalkar, Kaustubh Mhaisekar, Amaan Naikwadi, Archana Ghotkar
In recent years, there has been a considerable amount of research in the Gesture Recognition domain, mainly owing to the technological advancements in Computer Vision. Various new applications have been conceptualised and developed in this field. This paper discusses the implementation of gesture recognition systems in multiple sectors such as gaming, healthcare, home appliances, industrial robots, and virtual reality. Different methodologies for capturing gestures are compared and contrasted throughout this survey. Various data sources and data acquisition techniques have been discussed. The role of gestures in sign language has been studied and existing approaches have been reviewed. Common challenges faced while building gesture recognition systems have also been explored.
Authors: Quoc Hung T. Le, Hieu H. Pham
Musculoskeletal diseases and cognitive impairments in patients lead to difficulties in movement as well as negative effects on their psychological health. Clinical gait analysis, a vital tool for early diagnosis and treatment, traditionally relies on expensive optical motion capture systems. Recent advances in computer vision and deep learning have opened the door to more accessible and cost-effective alternatives. This paper introduces a novel spatio-temporal Transformer network to estimate critical gait parameters from RGB videos captured by a single-view camera. Empirical evaluations on a public dataset of cerebral palsy patients indicate that the proposed framework surpasses current state-of-the-art approaches and show significant improvements in predicting general gait parameters (including Walking Speed, Gait Deviation Index - GDI, and Knee Flexion Angle at Maximum Extension), while utilizing fewer parameters and alleviating the need for manual feature extraction.
Authors: Yaoyao Zhong, Mengshi Qi, Rui Wang, Yuhan Qiu, Yang Zhang, Huadong Ma
Video Internet of Things (VIoT) has shown full potential in collecting an unprecedented volume of video data. Learning to schedule perceiving models and analyzing the collected videos intelligently will be potential sparks for VIoT. In this paper, to address the challenges posed by the fine-grained and interrelated vision tool usage of VIoT, we build VIoTGPT, the framework based on LLMs to correctly interact with humans, query knowledge videos, and invoke vision models to accomplish complicated tasks. To support VIoTGPT and related future works, we meticulously crafted the training dataset and established benchmarks involving 11 representative vision models across three categories based on semi-automatic annotations. To guide LLM to act as the intelligent agent towards intelligent VIoT, we resort to ReAct instruction tuning based on the collected VIoT dataset to learn the tool capability. Quantitative and qualitative experimental results and analyses demonstrate the effectiveness of VIoTGPT.
Authors: Deepak Sridhar, Yunsheng Li, Nuno Vasconcelos
Vision Transformers have received significant attention due to their impressive performance in many vision tasks. While the token mixer or attention block has been studied in great detail, the channel mixer or feature mixing block (FFN or MLP) has not been explored in depth albeit it accounts for a bulk of the parameters and computation in a model. In this work, we study whether sparse feature mixing can replace the dense connections and confirm this with a block diagonal MLP structure that improves the accuracy by supporting larger expansion ratios. To improve the feature clusters formed by this structure and thereby further improve the accuracy, a lightweight, parameter-free, channel covariance attention (CCA) mechanism is introduced as a parallel branch during training. This design of CCA enables gradual feature mixing across channel groups during training whose contribution decays to zero as the training progresses to convergence. This allows the CCA block to be discarded during inference, thus enabling enhanced performance with no additional computational cost. The resulting $\textit{Scalable CHannEl MixEr}$ (SCHEME) can be plugged into any ViT architecture to obtain a gamut of models with different trade-offs between complexity and performance by controlling the block diagonal structure size in the MLP. This is shown by the introduction of a new family of SCHEMEformer models. Experiments on image classification, object detection, and semantic segmentation, with different ViT backbones, consistently demonstrate substantial accuracy gains over existing designs, especially under lower FLOPs regimes. For example, the SCHEMEformer establishes a new SOTA of 79.7% accuracy for ViTs using pure attention mixers on ImageNet-1K at 1.77G FLOPs.
Authors: Taichi Nishimura, Shota Nakada, Masayoshi Kondo
In this paper, we propose an efficient and high-performance method for partially relevant video retrieval (PRVR), which aims to retrieve untrimmed long videos that contain at least one relevant moment to the input text query. In terms of both efficiency and performance, the overlooked bottleneck of previous studies is the visual encoding of dense frames. This guides researchers to choose lightweight visual backbones, yielding sub-optimal retrieval performance due to their limited capabilities of learned visual representations. However, it is undesirable to simply replace them with high-performance large-scale vision-and-language models (VLMs) due to their low efficiency. To address these issues, instead of dense frames, we focus on super images, which are created by rearranging the video frames in a $N \times N$ grid layout. This reduces the number of visual encodings to $\frac{1}{N^2}$ and compensates for the low efficiency of large-scale VLMs, allowing us to adopt them as powerful encoders. Surprisingly, we discover that with a simple query-image attention trick, VLMs generalize well to super images effectively and demonstrate promising zero-shot performance against SOTA methods efficiently. In addition, we propose a fine-tuning approach by incorporating a few trainable modules into the VLM backbones. The experimental results demonstrate that our approaches efficiently achieve the best performance on ActivityNet Captions and TVR.
Authors: Hamid Sarmadi, Thorsteinn Rögnvaldsson, Nils Roger Carlsson, Mattias Ohlsson, Ibrahim Wahab, Ola Hall
Deep convolutional neural networks (CNNs) have been shown to predict poverty and development indicators from satellite images with surprising accuracy. This paper presents a first attempt at analyzing the CNNs responses in detail and explaining the basis for the predictions. The CNN model, while trained on relatively low resolution day- and night-time satellite images, is able to outperform human subjects who look at high-resolution images in ranking the Wealth Index categories. Multiple explainability experiments performed on the model indicate the importance of the sizes of the objects, pixel colors in the image, and provide a visualization of the importance of different structures in input images. A visualization is also provided of type images that maximize the network prediction of Wealth Index, which provides clues on what the CNN prediction is based on.
Authors: Pietro Bonazzi, Sizhen Bian, Giovanni Lippolis, Yawei Li, Sadique Sheik, Michele Magno
This paper introduces a neuromorphic methodology for eye tracking, harnessing pure event data captured by a Dynamic Vision Sensor (DVS) camera. The framework integrates a directly trained Spiking Neuron Network (SNN) regression model and leverages a state-of-the-art low power edge neuromorphic processor - Speck, collectively aiming to advance the precision and efficiency of eye-tracking systems. First, we introduce a representative event-based eye-tracking dataset, "Ini-30", which was collected with two glass-mounted DVS cameras from thirty volunteers. Then,a SNN model, based on Integrate And Fire (IAF) neurons, named "Retina", is described , featuring only 64k parameters (6.63x fewer than the latest) and achieving pupil tracking error of only 3.24 pixels in a 64x64 DVS input. The continous regression output is obtained by means of convolution using a non-spiking temporal 1D filter slided across the output spiking layer. Finally, we evaluate Retina on the neuromorphic processor, showing an end-to-end power between 2.89-4.8 mW and a latency of 5.57-8.01 mS dependent on the time window. We also benchmark our model against the latest event-based eye-tracking method, "3ET", which was built upon event frames. Results show that Retina achieves superior precision with 1.24px less pupil centroid error and reduced computational complexity with 35 times fewer MAC operations. We hope this work will open avenues for further investigation of close-loop neuromorphic solutions and true event-based training pursuing edge performance.
Authors: Pooja Bhatnagar, Sai Mrunaal, Sachin Kamnure
This research explores the realm of neural image captioning using deep learning models. The study investigates the performance of different neural architecture configurations, focusing on the inject architecture, and proposes a novel quality metric for evaluating caption generation. Through extensive experimentation and analysis, this work sheds light on the challenges and opportunities in image captioning, providing insights into model behavior and overfitting. The results reveal that while the merge models exhibit a larger vocabulary and higher ROUGE scores, the inject architecture generates relevant and concise image captions. The study also highlights the importance of refining training data and optimizing hyperparameters for improved model performance. This research contributes to the growing body of knowledge in neural image captioning and encourages further exploration in the field, emphasizing the democratization of artificial intelligence.
Authors: Yingzi Ma, Yulong Cao, Jiachen Sun, Marco Pavone, Chaowei Xiao
The quest for fully autonomous vehicles (AVs) capable of navigating complex real-world scenarios with human-like understanding and responsiveness. In this paper, we introduce Dolphins, a novel vision-language model architected to imbibe human-like abilities as a conversational driving assistant. Dolphins is adept at processing multimodal inputs comprising video (or image) data, text instructions, and historical control signals to generate informed outputs corresponding to the provided instructions. Building upon the open-sourced pretrained Vision-Language Model, OpenFlamingo, we first enhance Dolphins's reasoning capabilities through an innovative Grounded Chain of Thought (GCoT) process. Then we tailored Dolphins to the driving domain by constructing driving-specific instruction data and conducting instruction tuning. Through the utilization of the BDD-X dataset, we designed and consolidated four distinct AV tasks into Dolphins to foster a holistic understanding of intricate driving scenarios. As a result, the distinctive features of Dolphins are characterized into two dimensions: (1) the ability to provide a comprehensive understanding of complex and long-tailed open-world driving scenarios and solve a spectrum of AV tasks, and (2) the emergence of human-like capabilities including gradient-free instant adaptation via in-context learning and error recovery via reflection.
Authors: Zehao Zhu, Zhiwen Fan, Yifan Jiang, Zhangyang Wang
Novel view synthesis from limited observations remains an important and persistent task. However, high efficiency in existing NeRF-based few-shot view synthesis is often compromised to obtain an accurate 3D representation. To address this challenge, we propose a few-shot view synthesis framework based on 3D Gaussian Splatting that enables real-time and photo-realistic view synthesis with as few as three training views. The proposed method, dubbed FSGS, handles the extremely sparse initialized SfM points with a thoughtfully designed Gaussian Unpooling process. Our method iteratively distributes new Gaussians around the most representative locations, subsequently infilling local details in vacant areas. We also integrate a large-scale pre-trained monocular depth estimator within the Gaussians optimization process, leveraging online augmented views to guide the geometric optimization towards an optimal solution. Starting from sparse points observed from limited input viewpoints, our FSGS can accurately grow into unseen regions, comprehensively covering the scene and boosting the rendering quality of novel views. Overall, FSGS achieves state-of-the-art performance in both accuracy and rendering efficiency across diverse datasets, including LLFF, Mip-NeRF360, and Blender. Project website: https://zehaozhu.github.io/FSGS/.
Authors: Yajie Liu, Pu Ge, Haoxiang Ma, Shichao Fan, Qingjie Liu, Di Huang, Yunhong Wang
Referring image segmentation (RIS) aims to segment objects in an image conditioning on free-from text descriptions. Despite the overwhelming progress, it still remains challenging for current approaches to perform well on cases with various text expressions or with unseen visual entities, limiting its further application. In this paper, we present a novel RIS approach, which substantially improves the generalization ability by addressing the two dilemmas mentioned above. Specially, to deal with unconstrained texts, we propose to boost a given expression with an explicit and crucial prompt, which complements the expression in a unified context, facilitating target capturing in the presence of linguistic style changes. Furthermore, we introduce a multi-modal fusion aggregation module with visual guidance from a powerful pretrained model to leverage spatial relations and pixel coherences to handle the incomplete target masks and false positive irregular clumps which often appear on unseen visual entities. Extensive experiments are conducted in the zero-shot cross-dataset settings and the proposed approach achieves consistent gains compared to the state-of-the-art, e.g., 4.15\%, 5.45\%, and 4.64\% mIoU increase on RefCOCO, RefCOCO+ and ReferIt respectively, demonstrating its effectiveness. Additionally, the results on GraspNet-RIS show that our approach also generalizes well to new scenarios with large domain shifts.
Authors: Laura Smets, Werner Van Leekwijck, Ing Jyh Tsang, Steven Latré
Hyperdimensional Computing (HDC) is a brain-inspired and light-weight machine learning method. It has received significant attention in the literature as a candidate to be applied in the wearable internet of things, near-sensor artificial intelligence applications and on-device processing. HDC is computationally less complex than traditional deep learning algorithms and typically achieves moderate to good classification performance. A key aspect that determines the performance of HDC is the encoding of the input data to the hyperdimensional (HD) space. This article proposes a novel light-weight approach relying only on native HD arithmetic vector operations to encode binarized images that preserves similarity of patterns at nearby locations by using point of interest selection and local linear mapping. The method reaches an accuracy of 97.35% on the test set for the MNIST data set and 84.12% for the Fashion-MNIST data set. These results outperform other studies using baseline HDC with different encoding approaches and are on par with more complex hybrid HDC models. The proposed encoding approach also demonstrates a higher robustness to noise and blur compared to the baseline encoding.
Authors: Kerui Gu, Zhihao Li, Shiyong Liu, Jianzhuang Liu, Songcen Xu, Youliang Yan, Michael Bi Mi, Kenji Kawaguchi, Angela Yao
Estimating 3D rotations is a common procedure for 3D computer vision. The accuracy depends heavily on the rotation representation. One form of representation -- rotation matrices -- is popular due to its continuity, especially for pose estimation tasks. The learning process usually incorporates orthogonalization to ensure orthonormal matrices. Our work reveals, through gradient analysis, that common orthogonalization procedures based on the Gram-Schmidt process and singular value decomposition will slow down training efficiency. To this end, we advocate removing orthogonalization from the learning process and learning unorthogonalized `Pseudo' Rotation Matrices (PRoM). An optimization analysis shows that PRoM converges faster and to a better solution. By replacing the orthogonalization incorporated representation with our proposed PRoM in various rotation-related tasks, we achieve state-of-the-art results on large-scale benchmarks for human pose estimation.
Authors: A.M. Ershov, D.V. Tropin, E.E. Limonova, D.P. Nikolaev, V.V. Arlazarov
Presentation of folded documents is not an uncommon case in modern society. Digitizing such documents by capturing them with a smartphone camera can be tricky since a crease can divide the document contents into separate planes. To unfold the document, one could hold the edges potentially obscuring it in a captured image. While there are many geometrical rectification methods, they were usually developed for arbitrary bends and folds. We consider such algorithms and propose a novel approach Unfolder developed specifically for images of documents with a crease from folding in half. Unfolder is robust to projective distortions of the document image and does not fragment the image in the vicinity of a crease after rectification. A new Folded Document Images dataset was created to investigate the rectification accuracy of folded (2, 3, 4, and 8 folds) documents. The dataset includes 1600 images captured when document placed on a table and when held in hand. The Unfolder algorithm allowed for a recognition error rate of 0.33, which is better than the advanced neural network methods DocTr (0.44) and DewarpNet (0.57). The average runtime for Unfolder was only 0.25 s/image on an iPhone XR.
Authors: Wahidul Hasan Abir, Md. Fahim Uddin, Faria Rahman Khanam, Mohammad Monirujjaman Khan
This research paper focuses on Acute Lymphoblastic Leukemia (ALL), a form of blood cancer prevalent in children and teenagers, characterized by the rapid proliferation of immature white blood cells (WBCs). These atypical cells can overwhelm healthy cells, leading to severe health consequences. Early and accurate detection of ALL is vital for effective treatment and improving survival rates. Traditional diagnostic methods are time-consuming, costly, and prone to errors. The paper proposes an automated detection approach using computer-aided diagnostic (CAD) models, leveraging deep learning techniques to enhance the accuracy and efficiency of leukemia diagnosis. The study utilizes various transfer learning models like ResNet101V2, VGG19, InceptionV3, and InceptionResNetV2 for classifying ALL. The methodology includes using the Local Interpretable Model-Agnostic Explanations (LIME) for ensuring the validity and reliability of the AI system's predictions. This approach is critical for overcoming the "black box" nature of AI, where decisions made by models are often opaque and unaccountable. The paper highlights that the proposed method using the InceptionV3 model achieved an impressive 98.38% accuracy, outperforming other tested models. The results, verified by the LIME algorithm, showcase the potential of this method in accurately identifying ALL, providing a valuable tool for medical practitioners. The research underscores the impact of explainable artificial intelligence (XAI) in medical diagnostics, paving the way for more transparent and trustworthy AI applications in healthcare.
Authors: Mohammad Altillawi, Zador Pataki, Shile Li, Ziyuan Liu
Re-localizing a camera from a single image in a previously mapped area is vital for many computer vision applications in robotics and augmented/virtual reality. In this work, we address the problem of estimating the 6 DoF camera pose relative to a global frame from a single image. We propose to leverage a novel network of relative spatial and temporal geometric constraints to guide the training of a Deep Network for localization. We employ simultaneously spatial and temporal relative pose constraints that are obtained not only from adjacent camera frames but also from camera frames that are distant in the spatio-temporal space of the scene. We show that our method, through these constraints, is capable of learning to localize when little or very sparse ground-truth 3D coordinates are available. In our experiments, this is less than 1% of available ground-truth data. We evaluate our method on 3 common visual localization datasets and show that it outperforms other direct pose estimation methods.
Authors: Agnieszka Cisek, Karolina Korycinska, Leszek Pyziak, Marzena Malicka, Tomasz Wiecek, Grzegorz Gruzel, Kamil Szmuc, Jozef Cebulski, Mariusz Spyra
Diabetic retinopathy (DR) is a growing health problem worldwide and is a leading cause of visual impairment and blindness, especially among working people aged 20-65. Its incidence is increasing along with the number of diabetes cases, and it is more common in developed countries than in developing countries. Recent research in the field of diabetic retinopathy diagnosis is using advanced technologies, such as analysis of images obtained by ophthalmoscopy. Automatic methods for analyzing eye images based on neural networks, deep learning and image analysis algorithms can improve the efficiency of diagnosis. This paper describes an automatic DR diagnosis method that includes processing and analysis of ophthalmoscopic images of the eye. It uses morphological algorithms to identify the optic disc and lesions characteristic of DR, such as microaneurysms, hemorrhages and exudates. Automated DR diagnosis has the potential to improve the efficiency of early detection of this disease and contribute to reducing the number of cases of diabetes-related visual impairment. The final step was to create an application with a graphical user interface that allowed retinal images taken at cooperating ophthalmology offices to be uploaded to the server. These images were then analyzed using a developed algorithm to make a diagnosis.
Authors: Christina Gsaxner, Shohei Mori, Dieter Schmalstieg, Jan Egger, Gerhard Paar, Werner Bailer, Denis Kalkofen
Diminished reality (DR) refers to the removal of real objects from the environment by virtually replacing them with their background. Modern DR frameworks use inpainting to hallucinate unobserved regions. While recent deep learning-based inpainting is promising, the DR use case is complicated by the need to generate coherent structure and 3D geometry (i.e., depth), in particular for advanced applications, such as 3D scene editing. In this paper, we propose DeepDR, a first RGB-D inpainting framework fulfilling all requirements of DR: Plausible image and geometry inpainting with coherent structure, running at real-time frame rates, with minimal temporal artifacts. Our structure-aware generative network allows us to explicitly condition color and depth outputs on the scene semantics, overcoming the difficulty of reconstructing sharp and consistent boundaries in regions with complex backgrounds. Experimental results show that the proposed framework can outperform related work qualitatively and quantitatively.
Authors: Jose Luis Apellániz, Mikel García, Nerea Aranjuelo, Javier Barandiarán, Marcos Nieto
Curb detection is essential for environmental awareness in Automated Driving (AD), as it typically limits drivable and non-drivable areas. Annotated data are necessary for developing and validating an AD function. However, the number of public datasets with annotated point cloud curbs is scarce. This paper presents a method for detecting 3D curbs in a sequence of point clouds captured from a LiDAR sensor, which consists of two main steps. First, our approach detects the curbs at each scan using a segmentation deep neural network. Then, a sequence-level processing step estimates the 3D curbs in the reconstructed point cloud using the odometry of the vehicle. From these 3D points of the curb, we obtain polylines structured following ASAM OpenLABEL standard. These detections can be used as pre-annotations in labelling pipelines to efficiently generate curb-related ground truth data. We validate our approach through an experiment in which different human annotators were required to annotate curbs in a group of LiDAR-based sequences with and without our automatically generated pre-annotations. The results show that the manual annotation time is reduced by 50.99% thanks to our detections, keeping the data quality level.
Authors: Sungho Choi, Seungyul Han, Woojun Kim, Jongseong Chae, Whiyoung Jung, Youngchul Sung
In this paper, we consider domain-adaptive imitation learning with visual observation, where an agent in a target domain learns to perform a task by observing expert demonstrations in a source domain. Domain adaptive imitation learning arises in practical scenarios where a robot, receiving visual sensory data, needs to mimic movements by visually observing other robots from different angles or observing robots of different shapes. To overcome the domain shift in cross-domain imitation learning with visual observation, we propose a novel framework for extracting domain-independent behavioral features from input observations that can be used to train the learner, based on dual feature extraction and image reconstruction. Empirical results demonstrate that our approach outperforms previous algorithms for imitation learning from visual observation with domain shift.
Authors: Aleksi Knuutila
This paper presents a novel application of Generative Adverserial Networks (GANs) to study visual aspects of social processes. I train a a StyleGAN2-model on a custom dataset of 14,564 images of London, sourced from Google Streetview taken in London. After training, I invert the images in the training set, finding points in the model's latent space that correspond to them, and compare results from three inversion techniques. I connect each data point with metadata from the Indices of Multiple Deprivation, describing income, health and environmental quality in the area where the photographs were taken. It is then possible to map which parts of the model's latent space encode visual features that are distinctive for health, income and environmental quality, and condition the synthesis of new images based on these factors. The synthetic images created reflect visual features of social processes that were previously unknown and difficult to study, describing recurring visual differences between deprived and privileged areas in London. GANs are known for their capability to produce a continuous range of images that exhibit visual differences. The paper tests how to exploit this ability through visual comparisons in still images as well as through an interactive website where users can guide image synthesis with sliders. Though conditioned synthesis has its limitations and the results are difficult to validate, the paper points to the potential for generative models to be repurposed to be parts of social scientific methods.
Authors: Bardienus P. Duisterhof, Zhao Mandi, Yunchao Yao, Jia-Wei Liu, Mike Zheng Shou, Shuran Song, Jeffrey Ichnowski
Accurate 3D tracking in highly deformable scenes with occlusions and shadows can facilitate new applications in robotics, augmented reality, and generative AI. However, tracking under these conditions is extremely challenging due to the ambiguity that arises with large deformations, shadows, and occlusions. We introduce MD-Splatting, an approach for simultaneous 3D tracking and novel view synthesis, using video captures of a dynamic scene from various camera poses. MD-Splatting builds on recent advances in Gaussian splatting, a method that learns the properties of a large number of Gaussians for state-of-the-art and fast novel view synthesis. MD-Splatting learns a deformation function to project a set of Gaussians with non-metric, thus canonical, properties into metric space. The deformation function uses a neural-voxel encoding and a multilayer perceptron (MLP) to infer Gaussian position, rotation, and a shadow scalar. We enforce physics-inspired regularization terms based on local rigidity, conservation of momentum, and isometry, which leads to trajectories with smaller trajectory errors. MD-Splatting achieves high-quality 3D tracking on highly deformable scenes with shadows and occlusions. Compared to state-of-the-art, we improve 3D tracking by an average of 23.9 %, while simultaneously achieving high-quality novel view synthesis. With sufficient texture such as in scene 6, MD-Splatting achieves a median tracking error of 3.39 mm on a cloth of 1 x 1 meters in size. Project website: https://md-splatting.github.io/.
Authors: Zhaoning Wang, Ming Li, Chen Chen
With the recent development of generative models, Text-to-3D generations have also seen significant growth. Nonetheless, achieving precise control over 3D generation continues to be an arduous task, as using text to control often leads to missing objects and imprecise locations. Contemporary strategies for enhancing controllability in 3D generation often entail the introduction of additional parameters, such as customized diffusion models. This often induces hardness in adapting to different diffusion models or creating distinct objects.
In this paper, we present LucidDreaming as an effective pipeline capable of fine-grained control over 3D generation. It requires only minimal input of 3D bounding boxes, which can be deduced from a simple text prompt using a Large Language Model. Specifically, we propose clipped ray sampling to separately render and optimize objects with user specifications. We also introduce object-centric density blob bias, fostering the separation of generated objects. With individual rendering and optimizing of objects, our method excels not only in controlled content generation from scratch but also within the pre-trained NeRF scenes. In such scenarios, existing generative approaches often disrupt the integrity of the original scene, and current editing methods struggle to synthesize new content in empty spaces. We show that our method exhibits remarkable adaptability across a spectrum of mainstream Score Distillation Sampling-based 3D generation frameworks, and achieves superior alignment of 3D content when compared to baseline approaches. We also provide a dataset of prompts with 3D bounding boxes, benchmarking 3D spatial controllability.
Authors: En Yu, Liang Zhao, Yana Wei, Jinrong Yang, Dongming Wu, Lingyu Kong, Haoran Wei, Tiancai Wang, Zheng Ge, Xiangyu Zhang, Wenbing Tao
Humans possess the remarkable ability to foresee the future to a certain extent based on present observations, a skill we term as foresight minds. However, this capability remains largely under explored within existing Multimodal Large Language Models (MLLMs), hindering their capacity to learn the fundamental principles of how things operate and the intentions behind the observed subjects. To address this issue, we introduce the integration of future modeling into the existing learning frameworks of MLLMs. By utilizing the subject trajectory, a highly structured representation of a consecutive frame sequence, as a learning objective, we aim to bridge the gap between the past and the future. We propose two innovative methods to empower MLLMs with foresight minds, Foresight Pre-Training (FPT) and Foresight Instruction-Tuning (FIT), which are inspired by the modern learning paradigm of LLMs. Specifically, FPT jointly training various tasks centered on trajectories, enabling MLLMs to learn how to attend and predict entire trajectories from a given initial observation. Then, FIT requires MLLMs to first predict trajectories of related objects and then reason about potential future events based on them. Aided by FPT and FIT, we build a novel and unified MLLM named Merlin that supports multi-images input and analysis about potential actions of multiple objects for the future reasoning. Experimental results show Merlin powerful foresight minds with impressive performance on both future reasoning and visual comprehension tasks.
Authors: Xudong Li, Jingyuan Zheng, Xiawu Zheng, Runze Hu, Enwei Zhang, Yuting Gao, Yunhang Shen, Ke Li, Yutao Liu, Pingyang Dai, Yan Zhang, Rongrong Ji
Image Quality Assessment (IQA) with reference images have achieved great success by imitating the human vision system, in which the image quality is effectively assessed by comparing the query image with its pristine reference image. However, for the images in the wild, it is quite difficult to access accurate reference images. We argue that it is possible to learn reference knowledge under the No-Reference Image Quality Assessment (NR-IQA) setting, which is effective and efficient empirically. Concretely, by innovatively introducing a novel feature distillation method in IQA, we propose a new framework to learn comparative knowledge from non-aligned reference images. And then, to achieve fast convergence and avoid overfitting, we further propose an inductive bias regularization. Such a framework not only solves the congenital defects of NR-IQA but also improves the feature extraction framework, enabling it to express more abundant quality information. Surprisingly, our method utilizes less input while obtaining a more significant improvement compared to the teacher models. Extensive experiments on eight standard NR-IQA datasets demonstrate the superior performance to the state-of-the-art NR-IQA methods, i.e., achieving the PLCC values of 0.917 (vs. 0.884 in LIVEC) and 0.686 (vs. 0.661 in LIVEFB).
Authors: Emma Cramer, Jonas Reiher, Sebastian Trimpe
Reinforcement learning (RL) for robot control typically requires a detailed representation of the environment state, including information about task-relevant objects not directly measurable. Keypoint detectors, such as spatial autoencoders (SAEs), are a common approach to extracting a low-dimensional representation from high-dimensional image data. SAEs aim at spatial features such as object positions, which are often useful representations in robotic RL. However, whether an SAE is actually able to track objects in the scene and thus yields a spatial state representation well suited for RL tasks has rarely been examined due to a lack of established metrics. In this paper, we propose to assess the performance of an SAE instance by measuring how well keypoints track ground truth objects in images. We present a computationally lightweight metric and use it to evaluate common baseline SAE architectures on image data from a simulated robot task. We find that common SAEs differ substantially in their spatial extraction capability. Furthermore, we validate that SAEs that perform well in our metric achieve superior performance when used in downstream RL. Thus, our metric is an effective and lightweight indicator of RL performance before executing expensive RL training. Building on these insights, we identify three key modifications of SAE architectures to improve tracking performance. We make our code available at anonymous.4open.science/r/sae-rl.
Authors: Sahar Nasirihaghighi, Negin Ghamsarian, Heinrich Husslein, Klaus Schoeffmann
Analyzing laparoscopic surgery videos presents a complex and multifaceted challenge, with applications including surgical training, intra-operative surgical complication prediction, and post-operative surgical assessment. Identifying crucial events within these videos is a significant prerequisite in a majority of these applications. In this paper, we introduce a comprehensive dataset tailored for relevant event recognition in laparoscopic gynecology videos. Our dataset includes annotations for critical events associated with major intra-operative challenges and post-operative complications. To validate the precision of our annotations, we assess event recognition performance using several CNN-RNN architectures. Furthermore, we introduce and evaluate a hybrid transformer architecture coupled with a customized training-inference framework to recognize four specific events in laparoscopic surgery videos. Leveraging the Transformer networks, our proposed architecture harnesses inter-frame dependencies to counteract the adverse effects of relevant content occlusion, motion blur, and surgical scene variation, thus significantly enhancing event recognition accuracy. Moreover, we present a frame sampling strategy designed to manage variations in surgical scenes and the surgeons' skill level, resulting in event recognition with high temporal resolution. We empirically demonstrate the superiority of our proposed methodology in event recognition compared to conventional CNN-RNN architectures through a series of extensive experiments.
Authors: Afifa Khaled, Chao Li, Jia Ning, Kun He
Normalization techniques have been widely used in the field of deep learning due to their capability of enabling higher learning rates and are less careful in initialization. However, the effectiveness of popular normalization technologies is typically limited to specific areas. Unlike the standard Batch Normalization (BN) and Layer Normalization (LN), where BN computes the mean and variance along the (N,H,W) dimensions and LN computes the mean and variance along the (C,H,W) dimensions (N, C, H and W are the batch, channel, spatial height and width dimension, respectively), this paper presents a novel normalization technique called Batch Channel Normalization (BCN). To exploit both the channel and batch dependence and adaptively and combine the advantages of BN and LN based on specific datasets or tasks, BCN separately normalizes inputs along the (N, H, W) and (C, H, W) axes, then combines the normalized outputs based on adaptive parameters. As a basic block, BCN can be easily integrated into existing models for various applications in the field of computer vision. Empirical results show that the proposed technique can be seamlessly applied to various versions of CNN or Vision Transformer architecture. The code is publicly available at https://github.com/AfifaKhaled/BatchChannel-Normalization
Authors: Md. Mahmudur Rahman, Sajid Islam, Showren Chowdhury, Sadia Jahan Zeba, Debajyoti Karmaker
This study delves into the flight behaviors of Budgerigars (Melopsittacus undulatus) to gain insights into their flight trajectories and movements. Using 3D reconstruction from stereo video camera recordings, we closely examine the velocity and acceleration patterns during three flight motion takeoff, flying and landing. The findings not only contribute to our understanding of bird behaviors but also hold significant implications for the advancement of algorithms in Unmanned Aerial Vehicles (UAVs). The research aims to bridge the gap between biological principles observed in birds and the application of these insights in developing more efficient and autonomous UAVs. In the context of the increasing use of drones, this study focuses on the biologically inspired principles drawn from bird behaviors, particularly during takeoff, flying and landing flight, to enhance UAV capabilities. The dataset created for this research sheds light on Budgerigars' takeoff, flying, and landing techniques, emphasizing their ability to control speed across different situations and surfaces. The study underscores the potential of incorporating these principles into UAV algorithms, addressing challenges related to short-range navigation, takeoff, flying, and landing.
Authors: João Carreira, Michael King, Viorica Pătrăucean, Dilara Gokay, Cătălin Ionescu, Yi Yang, Daniel Zoran, Joseph Heyward, Carl Doersch, Yusuf Aytar, Dima Damen, Andrew Zisserman
We introduce a framework for online learning from a single continuous video stream -- the way people and animals learn, without mini-batches, data augmentation or shuffling. This poses great challenges given the high correlation between consecutive video frames and there is very little prior work on it. Our framework allows us to do a first deep dive into the topic and includes a collection of streams and tasks composed from two existing video datasets, plus methodology for performance evaluation that considers both adaptation and generalization. We employ pixel-to-pixel modelling as a practical and flexible way to switch between pre-training and single-stream evaluation as well as between arbitrary tasks, without ever requiring changes to models and always using the same pixel loss. Equipped with this framework we obtained large single-stream learning gains from pre-training with a novel family of future prediction tasks, found that momentum hurts, and that the pace of weight updates matters. The combination of these insights leads to matching the performance of IID learning with batch size 1, when using the same architecture and without costly replay buffers.
Authors: Michail Tarasiou, Jiankang Deng, Stefanos Zafeiriou
Heterogeneous face recognition (HFR) involves the intricate task of matching face images across the visual domains of visible (VIS) and near-infrared (NIR). While much of the existing literature on HFR identifies the domain gap as a primary challenge and directs efforts towards bridging it at either the input or feature level, our work deviates from this trend. We observe that large neural networks, unlike their smaller counterparts, when pre-trained on large scale homogeneous VIS data, demonstrate exceptional zero-shot performance in HFR, suggesting that the domain gap might be less pronounced than previously believed. By approaching the HFR problem as one of low-data fine-tuning, we introduce a straightforward framework: comprehensive pre-training, succeeded by a regularized fine-tuning strategy, that matches or surpasses the current state-of-the-art on four publicly available benchmarks. Corresponding codes can be found at https://github.com/michaeltrs/RethinkNIRVIS.
Authors: Yuxin Li, Qiang Han, Mengying Yu, Yuxin Jiang, Chaikiat Yeo, Yiheng Li, Zihang Huang, Nini Liu, Hsuanhan Chen, Xiaojun Wu
3D object detection in Bird's-Eye-View (BEV) space has recently emerged as a prevalent approach in the field of autonomous driving. Despite the demonstrated improvements in accuracy and velocity estimation compared to perspective view methods, the deployment of BEV-based techniques in real-world autonomous vehicles remains challenging. This is primarily due to their reliance on vision-transformer (ViT) based architectures, which introduce quadratic complexity with respect to the input resolution. To address this issue, we propose an efficient BEV-based 3D detection framework called BEVENet, which leverages a convolutional-only architectural design to circumvent the limitations of ViT models while maintaining the effectiveness of BEV-based methods. Our experiments show that BEVENet is 3$\times$ faster than contemporary state-of-the-art (SOTA) approaches on the NuScenes challenge, achieving a mean average precision (mAP) of 0.456 and a nuScenes detection score (NDS) of 0.555 on the NuScenes validation dataset, with an inference speed of 47.6 frames per second. To the best of our knowledge, this study stands as the first to achieve such significant efficiency improvements for BEV-based methods, highlighting their enhanced feasibility for real-world autonomous driving applications.
Authors: Asifullah Khan, Zunaira Rauf, Abdul Rehman Khan, Saima Rathore, Saddam Hussain Khan, Sahar Shah, Umair Farooq, Hifsa Asif, Aqsa Asif, Umme Zahoora, Rafi Ullah Khalil, Suleman Qamar, Umme Hani Asif, Faiza Babar Khan, Abdul Majid, Jeonghwan Gwak
Medical image segmentation plays a crucial role in various healthcare applications, enabling accurate diagnosis, treatment planning, and disease monitoring. In recent years, Vision Transformers (ViTs) have emerged as a promising technique for addressing the challenges in medical image segmentation. In medical images, structures are usually highly interconnected and globally distributed. ViTs utilize their multi-scale attention mechanism to model the long-range relationships in the images. However, they do lack image-related inductive bias and translational invariance, potentially impacting their performance. Recently, researchers have come up with various ViT-based approaches that incorporate CNNs in their architectures, known as Hybrid Vision Transformers (HVTs) to capture local correlation in addition to the global information in the images. This survey paper provides a detailed review of the recent advancements in ViTs and HVTs for medical image segmentation. Along with the categorization of ViT and HVT-based medical image segmentation approaches we also present a detailed overview of their real-time applications in several medical image modalities. This survey may serve as a valuable resource for researchers, healthcare practitioners, and students in understanding the state-of-the-art approaches for ViT-based medical image segmentation.
Authors: Karim Kassab, Antoine Schnepf, Jean-Yves Franceschi, Laurent Caraffa, Jeremie Mary, Valérie Gouet-Brunet
Modeling large-scale scenes from unconstrained image collections in-the-wild has proven to be a major challenge in computer vision. Existing methods tackling in-the-wild neural rendering operate in a closed-world setting, where knowledge is limited to a scene's captured images within a training set. We propose EvE, which is, to the best of our knowledge, the first method leveraging generative priors to improve in-the-wild scene modeling. We employ pre-trained generative networks to enrich K-Planes representations with extrinsic knowledge. To this end, we define an alternating training procedure to conduct optimization guidance of K-Planes trained on the training set. We carry out extensive experiments and verify the merit of our method on synthetic data as well as real tourism photo collections. EvE enhances rendered scenes with richer details and outperforms the state of the art on the task of novel view synthesis in-the-wild. Our project page can be found at https://eve-nvs.github.io .
Authors: Ioannis Kakogeorgiou, Spyros Gidaris, Konstantinos Karantzalos, Nikos Komodakis
Unsupervised object-centric learning aims to decompose scenes into interpretable object entities, termed slots. Slot-based auto-encoders stand out as a prominent method for this task. Within them, crucial aspects include guiding the encoder to generate object-specific slots and ensuring the decoder utilizes them during reconstruction. This work introduces two novel techniques, (i) an attention-based self-training approach, which distills superior slot-based attention masks from the decoder to the encoder, enhancing object segmentation, and (ii) an innovative patch-order permutation strategy for autoregressive transformers that strengthens the role of slot vectors in reconstruction. The effectiveness of these strategies is showcased experimentally. The combined approach significantly surpasses prior slot-based autoencoder methods in unsupervised object segmentation, especially with complex real-world images. We provide the implementation code at https://github.com/gkakogeorgiou/spot .
Authors: Pengxiang Li, Zhili Liu, Kai Chen, Lanqing Hong, Yunzhi Zhuge, Dit-Yan Yeung, Huchuan Lu, Xu Jia
Diffusion models have gained prominence in generating data for perception tasks such as image classification and object detection. However, the potential in generating high-quality tracking sequences, a crucial aspect in the field of video perception, has not been fully investigated. To address this gap, we propose TrackDiffusion, a novel architecture designed to generate continuous video sequences from the tracklets. TrackDiffusion represents a significant departure from the traditional layout-to-image (L2I) generation and copy-paste synthesis focusing on static image elements like bounding boxes by empowering image diffusion models to encompass dynamic and continuous tracking trajectories, thereby capturing complex motion nuances and ensuring instance consistency among video frames. For the first time, we demonstrate that the generated video sequences can be utilized for training multi-object tracking (MOT) systems, leading to significant improvement in tracker performance. Experimental results show that our model significantly enhances instance consistency in generated video sequences, leading to improved perceptual metrics. Our approach achieves an improvement of 8.7 in TrackAP and 11.8 in TrackAP$_{50}$ on the YTVIS dataset, underscoring its potential to redefine the standards of video data generation for MOT tasks and beyond.
Authors: Junwei Yang, Pietro Liò
Purpose: To develop an efficient dual-domain reconstruction framework for multi-contrast MRI, with the focus on minimising cross-contrast misalignment in both the image and the frequency domains to enhance optimisation. Theory and Methods: Our proposed framework, based on deep learning, facilitates the optimisation for under-sampled target contrast using fully-sampled reference contrast that is quicker to acquire. The method consists of three key steps: 1) Learning to synthesise data resembling the target contrast from the reference contrast; 2) Registering the multi-contrast data to reduce inter-scan motion; and 3) Utilising the registered data for reconstructing the target contrast. These steps involve learning in both domains with regularisation applied to ensure their consistency. We also compare the reconstruction performance with existing deep learning-based methods using a dataset of brain MRI scans. Results: Extensive experiments demonstrate the superiority of our proposed framework, for up to an 8-fold acceleration rate, compared to state-of-the-art algorithms. Comprehensive analysis and ablation studies further present the effectiveness of the proposed components. Conclusion:Our dual-domain framework offers a promising approach to multi-contrast MRI reconstruction. It can also be integrated with existing methods to further enhance the reconstruction.
Authors: Kangcheng Liu, Yong-Jin Liu, Kai Tang, Ming Liu, Baoquan Chen
Deep neural network models have achieved remarkable progress in 3D scene understanding while trained in the closed-set setting and with full labels. However, the major bottleneck for current 3D recognition approaches is that they do not have the capacity to recognize any unseen novel classes beyond the training categories in diverse kinds of real-world applications. In the meantime, current state-of-the-art 3D scene understanding approaches primarily require high-quality labels to train neural networks, which merely perform well in a fully supervised manner. This work presents a generalized and simple framework for dealing with 3D scene understanding when the labeled scenes are quite limited. To extract knowledge for novel categories from the pre-trained vision-language models, we propose a hierarchical feature-aligned pre-training and knowledge distillation strategy to extract and distill meaningful information from large-scale vision-language models, which helps benefit the open-vocabulary scene understanding tasks. To leverage the boundary information, we propose a novel energy-based loss with boundary awareness benefiting from the region-level boundary predictions. To encourage latent instance discrimination and to guarantee efficiency, we propose the unsupervised region-level semantic contrastive learning scheme for point clouds, using confident predictions of the neural network to discriminate the intermediate feature embeddings at multiple stages. Extensive experiments with both indoor and outdoor scenes demonstrated the effectiveness of our approach in both data-efficient learning and open-world few-shot learning. All codes, models, and data are made publicly available at: https://drive.google.com/drive/folders/1M58V-PtR8DBEwD296zJkNg_m2qq-MTAP?usp=sharing.
Authors: Mehdi Naouar, Gabriel Kalweit, Anusha Klett, Yannick Vogt, Paula Silvestrini, Diana Laura Infante Ramirez, Roland Mertelsmann, Joschka Boedecker, Maria Kalweit
In recent years, several unsupervised cell segmentation methods have been presented, trying to omit the requirement of laborious pixel-level annotations for the training of a cell segmentation model. Most if not all of these methods handle the instance segmentation task by focusing on the detection of different cell instances ignoring their type. While such models prove adequate for certain tasks, like cell counting, other applications require the identification of each cell's type. In this paper, we present CellMixer, an innovative annotation-free approach for the semantic segmentation of heterogeneous cell populations. Our augmentation-based method enables the training of a segmentation model from image-level labels of homogeneous cell populations. Our results show that CellMixer can achieve competitive segmentation performance across multiple cell types and imaging modalities, demonstrating the method's scalability and potential for broader applications in medical imaging, cellular biology, and diagnostics.
Authors: Ying Nie, Wei He, Kai Han, Yehui Tang, Tianyu Guo, Fanyi Du, Yunhe Wang
Vision-language pre-training like CLIP has shown promising performance on various downstream tasks such as zero-shot image classification and image-text retrieval. Most of the existing CLIP-alike works usually adopt relatively large image encoders like ResNet50 and ViT, while the lightweight counterparts are rarely discussed. In this paper, we propose a multi-level interaction paradigm for training lightweight CLIP models. Firstly, to mitigate the problem that some image-text pairs are not strictly one-to-one correspondence, we improve the conventional global instance-level alignment objective by softening the label of negative samples progressively. Secondly, a relaxed bipartite matching based token-level alignment objective is introduced for finer-grained alignment between image patches and textual words. Moreover, based on the observation that the accuracy of CLIP model does not increase correspondingly as the parameters of text encoder increase, an extra objective of masked language modeling (MLM) is leveraged for maximizing the potential of the shortened text encoder. In practice, an auxiliary fusion module injecting unmasked image embedding into masked text embedding at different network stages is proposed for enhancing the MLM. Extensive experiments show that without introducing additional computational cost during inference, the proposed method achieves a higher performance on multiple downstream tasks.
Authors: Junwei Yang, Pietro Liò
In recent studies on MRI reconstruction, advances have shown significant promise for further accelerating the MRI acquisition. Most state-of-the-art methods require a large amount of fully-sampled data to optimise reconstruction models, which is impractical and expensive under certain clinical settings. On the other hand, for unsupervised scan-specific reconstruction methods, overfitting is likely to happen due to insufficient supervision, while restrictions on acceleration rates and under-sampling patterns further limit their applicability. To this end, we propose an unsupervised, adaptive coarse-to-fine framework that enhances reconstruction quality without being constrained by the sparsity levels or patterns in under-sampling. The framework employs an implicit neural representation for scan-specific MRI reconstruction, learning a mapping from multi-dimensional coordinates to their corresponding signal intensities. Moreover, we integrate a novel learning strategy that progressively refines the use of acquired k-space signals for self-supervision. This approach effectively adjusts the proportion of supervising signals from unevenly distributed information across different frequency bands, thus mitigating the issue of overfitting while improving the overall reconstruction. Comprehensive evaluation on a public dataset, including both 2D and 3D data, has shown that our method outperforms current state-of-the-art scan-specific MRI reconstruction techniques, for up to 8-fold under-sampling.
Authors: Yongsong Huang, Shinichiro Omachi
The ability of generative models to accurately fit data distributions has resulted in their widespread adoption and success in fields such as computer vision and natural language processing. In this chapter, we provide a brief overview of the application of generative models in the domain of infrared (IR) image super-resolution, including a discussion of the various challenges and adversarial training methods employed. We propose potential areas for further investigation and advancement in the application of generative models for IR image super-resolution.
Authors: Jaime Corsetti, Davide Boscaini, Changjae Oh, Andrea Cavallaro, Fabio Poiesi
We introduce the new setting of open-vocabulary object 6D pose estimation, in which a textual prompt is used to specify the object of interest. In contrast to existing approaches, in our setting (i) the object of interest is specified solely through the textual prompt, (ii) no object model (e.g. CAD or video sequence) is required at inference, (iii) the object is imaged from two different viewpoints of two different scenes, and (iv) the object was not observed during the training phase. To operate in this setting, we introduce a novel approach that leverages a Vision-Language Model to segment the object of interest from two distinct scenes and to estimate its relative 6D pose. The key of our approach is a carefully devised strategy to fuse object-level information provided by the prompt with local image features, resulting in a feature space that can generalize to novel concepts. We validate our approach on a new benchmark based on two popular datasets, REAL275 and Toyota-Light, which collectively encompass 39 object instances appearing in four thousand image pairs. The results demonstrate that our approach outperforms both a well-established hand-crafted method and a recent deep learning-based baseline in estimating the relative 6D pose of objects in different scenes. Project website: https://jcorsetti.github.io/oryon-website/.
Authors: Benedikt W. Hosp, Martin Dechant, Yannick Sauer, Rajat Agarwala, Siegfried Wahl
Developing and evaluating vision science methods require robust and efficient tools for assessing their performance in various real-world scenarios. This study presents a novel virtual reality (VR) simulation tool that simulates real-world optical methods while giving high experimental control to the experiment. The tool incorporates an experiment controller, to smoothly and easily handle multiple conditions, a generic eye-tracking controller, that works with most common VR eye-trackers, a configurable defocus simulator, and a generic VR questionnaire loader to assess participants' behavior in virtual reality. This VR-based simulation tool bridges the gap between theoretical and applied research on new optical methods, corrections, and therapies. It enables vision scientists to increase their research tools with a robust, realistic, and fast research environment.
Authors: Martin Georg Ljungqvist, Otto Nordander, Markus Skans, Arvid Mildner, Tony Liu, Pierre Nugues
To train well-performing generalizing neural networks, sufficiently large and diverse datasets are needed. Collecting data while adhering to privacy legislation becomes increasingly difficult and annotating these large datasets is both a resource-heavy and time-consuming task. An approach to overcome these difficulties is to use synthetic data since it is inherently scalable and can be automatically annotated. However, how training on synthetic data affects the layers of a neural network is still unclear. In this paper, we train the YOLOv3 object detector on real and synthetic images from city environments. We perform a similarity analysis using Centered Kernel Alignment (CKA) to explore the effects of training on synthetic data on a layer-wise basis. The analysis captures the architecture of the detector while showing both different and similar patterns between different models. With this similarity analysis we want to give insights on how training synthetic data affects each layer and to give a better understanding of the inner workings of complex neural networks. The results show that the largest similarity between a detector trained on real data and a detector trained on synthetic data was in the early layers, and the largest difference was in the head part. The results also show that no major difference in performance or similarity could be seen between frozen and unfrozen backbone.
Authors: Bin Xiao, Murat Simsek, Burak Kantarci, Ala Abu Alkheir
Table Structure Recognition (TSR) aims at transforming unstructured table images into structured formats, such as HTML sequences. One type of popular solution is using detection models to detect components of a table, such as columns and rows, then applying a rule-based post-processing method to convert detection results into HTML sequences. However, existing detection-based studies often have the following limitations. First, these studies usually pay more attention to improving the detection performance, which does not necessarily lead to better performance regarding cell-level metrics, such as TEDS. Second, some solutions over-simplify the problem and can miss some critical information. Lastly, even though some studies defined the problem to detect more components to provide as much information as other types of solutions, these studies ignore the fact this problem definition is a multi-label detection because row, projected row header and column header can share identical bounding boxes. Besides, there is often a performance gap between two-stage and transformer-based detection models regarding the structure-only TEDS, even though they have similar performance regarding the COCO metrics. Therefore, we revisit the limitations of existing detection-based solutions, compare two-stage and transformer-based detection models, and identify the key design aspects for the success of a two-stage detection model for the TSR task, including the multi-class problem definition, the aspect ratio for anchor box generation, and the feature generation of the backbone network. We applied simple methods to improve these aspects of the Cascade R-CNN model, achieved state-of-the-art performance, and improved the baseline Cascade R-CNN model by 19.32%, 11.56% and 14.77% regarding the structure-only TEDS on SciTSR, FinTabNet, and PubTables1M datasets.
Authors: Chinmay Savadikar, Xi Song, Tianfu Wu
We present GIFT (Generative Interpretable Fine-tuning Transformers) for fine-tuning pretrained (often large) Transformer models at downstream tasks in a parameter-efficient way with built-in interpretability. Our GIFT is a deep parameter-residual learning method, which addresses two problems in fine-tuning a pretrained Transformer model: Where to apply the parameter-efficient fine-tuning (PEFT) to be extremely lightweight yet sufficiently expressive, and How to learn the PEFT to better exploit the knowledge of the pretrained model in a direct way? For the former, we select the final projection (linear) layer in the multi-head self-attention of a Transformer model, and verify its effectiveness. For the latter, in contrast to the prior art that directly introduce new model parameters (often in low-rank approximation form) to be learned in fine-tuning with downstream data, we propose a method for learning to generate the fine-tuning parameters. Our GIFT is a hyper-Transformer which take as input the pretrained parameters of the projection layer to generate its fine-tuning parameters using a proposed Parameter-to-Cluster Attention (PaCa). The PaCa results in a simple clustering-based forward explainer that plays the role of semantic segmentation in testing. In experiments, our proposed GIFT is tested on the VTAB benchmark and the fine-grained visual classification (FGVC) benchmark. It obtains significantly better performance than the prior art. Our code is available at https://github.com/savadikarc/gift
Authors: Loick Chambon, Eloi Zablocki, Mickael Chen, Florent Bartoccioni, Patrick Perez, Matthieu Cord
Bird's-eye View (BeV) representations have emerged as the de-facto shared space in driving applications, offering a unified space for sensor data fusion and supporting various downstream tasks. However, conventional models use grids with fixed resolution and range and face computational inefficiencies due to the uniform allocation of resources across all cells. To address this, we propose PointBeV, a novel sparse BeV segmentation model operating on sparse BeV cells instead of dense grids. This approach offers precise control over memory usage, enabling the use of long temporal contexts and accommodating memory-constrained platforms. PointBeV employs an efficient two-pass strategy for training, enabling focused computation on regions of interest. At inference time, it can be used with various memory/performance trade-offs and flexibly adjusts to new specific use cases. PointBeV achieves state-of-the-art results on the nuScenes dataset for vehicle, pedestrian, and lane segmentation, showcasing superior performance in static and temporal settings despite being trained solely with sparse signals. We will release our code along with two new efficient modules used in the architecture: Sparse Feature Pulling, designed for the effective extraction of features from images to BeV, and Submanifold Attention, which enables efficient temporal modeling. Our code is available at https://github.com/valeoai/PointBeV.
Authors: Mingqiao Ye, Martin Danelljan, Fisher Yu, Lei Ke
The recent Gaussian Splatting achieves high-quality and real-time novel-view synthesis of the 3D scenes. However, it is solely concentrated on the appearance and geometry modeling, while lacking in fine-grained object-level scene understanding. To address this issue, we propose Gaussian Grouping, which extends Gaussian Splatting to jointly reconstruct and segment anything in open-world 3D scenes. We augment each Gaussian with a compact Identity Encoding, allowing the Gaussians to be grouped according to their object instance or stuff membership in the 3D scene. Instead of resorting to expensive 3D labels, we supervise the Identity Encodings during the differentiable rendering by leveraging the 2D mask predictions by SAM, along with introduced 3D spatial consistency regularization. Comparing to the implicit NeRF representation, we show that the discrete and grouped 3D Gaussians can reconstruct, segment and edit anything in 3D with high visual quality, fine granularity and efficiency. Based on Gaussian Grouping, we further propose a local Gaussian Editing scheme, which shows efficacy in versatile scene editing applications, including 3D object removal, inpainting, colorization and scene recomposition. Our code and models will be at https://github.com/lkeab/gaussian-grouping.
Authors: Min Wei, Jingkai Zhou, Junyao Sun, Xuesong Zhang
Existing score distillation methods are sensitive to classifier-free guidance (CFG) scale: manifested as over-smoothness or instability at small CFG scales, while over-saturation at large ones. To explain and analyze these issues, we revisit the derivation of Score Distillation Sampling (SDS) and decipher existing score distillation with the Wasserstein Generative Adversarial Network (WGAN) paradigm. With the WGAN paradigm, we find that existing score distillation either employs a fixed sub-optimal discriminator or conducts incomplete discriminator optimization, resulting in the scale-sensitive issue. We propose the Adversarial Score Distillation (ASD), which maintains an optimizable discriminator and updates it using the complete optimization objective. Experiments show that the proposed ASD performs favorably in 2D distillation and text-to-3D tasks against existing methods. Furthermore, to explore the generalization ability of our WGAN paradigm, we extend ASD to the image editing task, which achieves competitive results. The project page and code are at https://github.com/2y7c3/ASD.
Authors: Jie Ren, Mingjie Li, Zexu Liu, Quanshi Zhang
This paper aims to define, quantify, and analyze the feature complexity that is learned by a DNN. We propose a generic definition for the feature complexity. Given the feature of a certain layer in the DNN, our method disentangles feature components of different complexity orders from the feature. We further design a set of metrics to evaluate the reliability, the effectiveness, and the significance of over-fitting of these feature components. Furthermore, we successfully discover a close relationship between the feature complexity and the performance of DNNs. As a generic mathematical tool, the feature complexity and the proposed metrics can also be used to analyze the success of network compression and knowledge distillation.
Authors: Xin Wang, Jie Ren, Shuyun Lin, Xiangming Zhu, Yisen Wang, Quanshi Zhang
In this paper, we use the interaction inside adversarial perturbations to explain and boost the adversarial transferability. We discover and prove the negative correlation between the adversarial transferability and the interaction inside adversarial perturbations. The negative correlation is further verified through different DNNs with various inputs. Moreover, this negative correlation can be regarded as a unified perspective to understand current transferability-boosting methods. To this end, we prove that some classic methods of enhancing the transferability essentially decease interactions inside adversarial perturbations. Based on this, we propose to directly penalize interactions during the attacking process, which significantly improves the adversarial transferability.
Authors: Kai Zhang, Yawei Li, Jingyun Liang, Jiezhang Cao, Yulun Zhang, Hao Tang, Deng-Ping Fan, Radu Timofte, Luc Van Gool
While recent years have witnessed a dramatic upsurge of exploiting deep neural networks toward solving image denoising, existing methods mostly rely on simple noise assumptions, such as additive white Gaussian noise (AWGN), JPEG compression noise and camera sensor noise, and a general-purpose blind denoising method for real images remains unsolved. In this paper, we attempt to solve this problem from the perspective of network architecture design and training data synthesis. Specifically, for the network architecture design, we propose a swin-conv block to incorporate the local modeling ability of residual convolutional layer and non-local modeling ability of swin transformer block, and then plug it as the main building block into the widely-used image-to-image translation UNet architecture. For the training data synthesis, we design a practical noise degradation model which takes into consideration different kinds of noise (including Gaussian, Poisson, speckle, JPEG compression, and processed camera sensor noises) and resizing, and also involves a random shuffle strategy and a double degradation strategy. Extensive experiments on AGWN removal and real image denoising demonstrate that the new network architecture design achieves state-of-the-art performance and the new degradation model can help to significantly improve the practicability. We believe our work can provide useful insights into current denoising research.
Authors: Kyle Gao, Yina Gao, Hongjie He, Dening Lu, Linlin Xu, Jonathan Li
Neural Radiance Field (NeRF) has recently become a significant development in the field of Computer Vision, allowing for implicit, neural network-based scene representation and novel view synthesis. NeRF models have found diverse applications in robotics, urban mapping, autonomous navigation, virtual reality/augmented reality, and more. Due to the growing popularity of NeRF and its expanding research area, we present a comprehensive survey of NeRF papers from the past two years. Our survey is organized into architecture and application-based taxonomies and provides an introduction to the theory of NeRF and its training via differentiable volume rendering. We also present a benchmark comparison of the performance and speed of key NeRF models. By creating this survey, we hope to introduce new researchers to NeRF, provide a helpful reference for influential works in this field, as well as motivate future research directions with our discussion section.
Authors: Ling Tang, Wen Shen, Zhanpeng Zhou, Yuefeng Chen, Quanshi Zhang
In this paper, we prove the representation defects of a cascaded convolutional decoder network, considering the capacity of representing different frequency components of an input sample. We conduct the discrete Fourier transform on each channel of the feature map in an intermediate layer of the decoder network. Then, we extend the 2D circular convolution theorem to represent the forward and backward propagations through convolutional layers in the frequency domain. Based on this, we prove three defects in representing feature spectrums. First, we prove that the convolution operation, the zero-padding operation, and a set of other settings all make a convolutional decoder network more likely to weaken high-frequency components. Second, we prove that the upsampling operation generates a feature spectrum, in which strong signals repetitively appear at certain frequencies. Third, we prove that if the frequency components in the input sample and frequency components in the target output for regression have a small shift, then the decoder usually cannot be effectively learned.
Authors: Mahboobeh Jafari, Afshin Shoeibi, Navid Ghassemi, Jonathan Heras, Sai Ho Ling, Amin Beheshti, Yu-Dong Zhang, Shui-Hua Wang, Roohallah Alizadehsani, Juan M. Gorriz, U. Rajendra Acharya, Hamid Alinejad Rokny
Myocarditis is a significant cardiovascular disease (CVD) that poses a threat to the health of many individuals by causing damage to the myocardium. The occurrence of microbes and viruses, including the likes of HIV, plays a crucial role in the development of myocarditis disease (MCD). The images produced during cardiac magnetic resonance imaging (CMRI) scans are low contrast, which can make it challenging to diagnose cardiovascular diseases. In other hand, checking numerous CMRI slices for each CVD patient can be a challenging task for medical doctors. To overcome the existing challenges, researchers have suggested the use of artificial intelligence (AI)-based computer-aided diagnosis systems (CADS). The presented paper outlines a CADS for the detection of MCD from CMR images, utilizing deep learning (DL) methods. The proposed CADS consists of several steps, including dataset, preprocessing, feature extraction, classification, and post-processing. First, the Z-Alizadeh dataset was selected for the experiments. Subsequently, the CMR images underwent various preprocessing steps, including denoising, resizing, as well as data augmentation (DA) via CutMix and MixUp techniques. In the following, the most current deep pre-trained and transformer models are used for feature extraction and classification on the CMR images. The findings of our study reveal that transformer models exhibit superior performance in detecting MCD as opposed to pre-trained architectures. In terms of DL architectures, the Turbulence Neural Transformer (TNT) model exhibited impressive accuracy, reaching 99.73% utilizing a 10-fold cross-validation approach. Additionally, to pinpoint areas of suspicion for MCD in CMRI images, the Explainable-based Grad Cam method was employed.
Authors: Xiaowei Hu, Min Shi, Weiyun Wang, Sitong Wu, Linjie Xing, Wenhai Wang, Xizhou Zhu, Lewei Lu, Jie Zhou, Xiaogang Wang, Yu Qiao, Jifeng Dai
Vision transformers have gained popularity recently, leading to the development of new vision backbones with improved features and consistent performance gains. However, these advancements are not solely attributable to novel feature transformation designs; certain benefits also arise from advanced network-level and block-level architectures. This paper aims to identify the real gains of popular convolution and attention operators through a detailed study. We find that the key difference among these feature transformation modules, such as attention or convolution, lies in their spatial feature aggregation approach, known as the "spatial token mixer" (STM). To facilitate an impartial comparison, we introduce a unified architecture to neutralize the impact of divergent network-level and block-level designs. Subsequently, various STMs are integrated into this unified framework for comprehensive comparative analysis. Our experiments on various tasks and an analysis of inductive bias show a significant performance boost due to advanced network-level and block-level designs, but performance differences persist among different STMs. Our detailed analysis also reveals various findings about different STMs, such as effective receptive fields and invariance tests. All models and codes used in this study are publicly available at \url{https://github.com/OpenGVLab/STM-Evaluation}.
Authors: Yanlong Yang, Jianan Liu, Tao Huang, Qing-Long Han, Gang Ma, Bing Zhu
In autonomous driving, LiDAR and radar play important roles in the perception of the surrounding environment. LiDAR provides accurate 3D spatial sensing information but cannot work in adverse weather like fog. On the other hand, the radar signal can be diffracted when encountering raindrops or mist particles thanks to its wavelength, but it suffers from large noise. Recent state-of-the-art works reveal that fusion of radar and LiDAR can lead to robust detection in adverse weather. The existing works adopt convolutional neural network architecture to extract features from each sensor data, then align and aggregate the two branch features to predict object detection results. However, these methods have low accuracy of bounding box estimations due to a simple design of label assignment and fusion strategies. In this paper, we propose a bird's-eye view fusion learning-based anchor box-free object detection system, which fuses the feature derived from the radar range-azimuth heatmap and the LiDAR point cloud to estimate possible objects. Different label assignment strategies have been designed to facilitate the consistency between the classification of foreground or background anchor points and the corresponding bounding box regressions. Furthermore, the performance of the proposed object detector is further enhanced by employing a novel interactive transformer module. The superior performance of the methods proposed in this paper has been demonstrated using the recently published Oxford Radar RobotCar dataset. Our system's average precision significantly outperforms the state-of-the-art method by 13.1% and 19.0% at IoU of 0.8 under 'Clear+Foggy' training conditions for 'Clear' and 'Foggy' testing, respectively.
Authors: Bao Gia Doan, Ehsan Abbasnejad, Javen Qinfeng Shi, Damith C. Ranasinghe
We present a new algorithm to learn a deep neural network model robust against adversarial attacks. Previous algorithms demonstrate an adversarially trained Bayesian Neural Network (BNN) provides improved robustness. We recognize the adversarial learning approach for approximating the multi-modal posterior distribution of a Bayesian model can lead to mode collapse; consequently, the model's achievements in robustness and performance are sub-optimal. Instead, we first propose preventing mode collapse to better approximate the multi-modal posterior distribution. Second, based on the intuition that a robust model should ignore perturbations and only consider the informative content of the input, we conceptualize and formulate an information gain objective to measure and force the information learned from both benign and adversarial training instances to be similar. Importantly. we prove and demonstrate that minimizing the information gain objective allows the adversarial risk to approach the conventional empirical risk. We believe our efforts provide a step toward a basis for a principled method of adversarially training BNNs. Our model demonstrate significantly improved robustness--up to 20%--compared with adversarial training and Adv-BNN under PGD attacks with 0.035 distortion on both CIFAR-10 and STL-10 datasets.
Authors: Youming Deng, Xueting Li, Sifei Liu, Ming-Hsuan Yang
We present a physics-based inverse rendering method that learns the illumination, geometry, and materials of a scene from posed multi-view RGB images. To model the illumination of a scene, existing inverse rendering works either completely ignore the indirect illumination or model it by coarse approximations, leading to sub-optimal illumination, geometry, and material prediction of the scene. In this work, we propose a physics-based illumination model that first locates surface points through an efficient refined sphere tracing algorithm, then explicitly traces the incoming indirect lights at each surface point based on reflection. Then, we estimate each identified indirect light through an efficient neural network. Moreover, we utilize the Leibniz's integral rule to resolve non-differentiability in the proposed illumination model caused by boundary lights inspired by differentiable irradiance in computer graphics. As a result, the proposed differentiable illumination model can be learned end-to-end together with geometry and materials estimation. As a side product, our physics-based inverse rendering model also facilitates flexible and realistic material editing as well as relighting. Extensive experiments on synthetic and real-world datasets demonstrate that the proposed method performs favorably against existing inverse rendering methods on novel view synthesis and inverse rendering.
Authors: HyunJun Jung, Guangyao Zhai, Shun-Cheng Wu, Patrick Ruhkamp, Hannah Schieber, Giulia Rizzoli, Pengyuan Wang, Hongcheng Zhao, Lorenzo Garattoni, Sven Meier, Daniel Roth, Nassir Navab, Benjamin Busam
Estimating 6D object poses is a major challenge in 3D computer vision. Building on successful instance-level approaches, research is shifting towards category-level pose estimation for practical applications. Current category-level datasets, however, fall short in annotation quality and pose variety. Addressing this, we introduce HouseCat6D, a new category-level 6D pose dataset. It features 1) multi-modality with Polarimetric RGB and Depth (RGBD+P), 2) encompasses 194 diverse objects across 10 household categories, including two photometrically challenging ones, and 3) provides high-quality pose annotations with an error range of only 1.35 mm to 1.74 mm. The dataset also includes 4) 41 large-scale scenes with comprehensive viewpoint and occlusion coverage, 5) a checkerboard-free environment, and 6) dense 6D parallel-jaw robotic grasp annotations. Additionally, we present benchmark results for leading category-level pose estimation networks.
Authors: Erik Wallin, Lennart Svensson, Fredrik Kahl, Lars Hammarstrand
Open-set semi-supervised learning (OSSL) embodies a practical scenario within semi-supervised learning, wherein the unlabeled training set encompasses classes absent from the labeled set. Many existing OSSL methods assume that these out-of-distribution data are harmful and put effort into excluding data belonging to unknown classes from the training objective. In contrast, we propose an OSSL framework that facilitates learning from all unlabeled data through self-supervision. Additionally, we utilize an energy-based score to accurately recognize data belonging to the known classes, making our method well-suited for handling uncurated data in deployment. We show through extensive experimental evaluations that our method yields state-of-the-art results on many of the evaluated benchmark problems in terms of closed-set accuracy and open-set recognition when compared with existing methods for OSSL. Our code is available at https://github.com/walline/ssl-tf2-sefoss.
Authors: Mengyun Qiao, Shuo Wang, Huaqi Qiu, Antonio de Marvao, Declan P. O'Regan, Daniel Rueckert, Wenjia Bai
Two key questions in cardiac image analysis are to assess the anatomy and motion of the heart from images; and to understand how they are associated with non-imaging clinical factors such as gender, age and diseases. While the first question can often be addressed by image segmentation and motion tracking algorithms, our capability to model and to answer the second question is still limited. In this work, we propose a novel conditional generative model to describe the 4D spatio-temporal anatomy of the heart and its interaction with non-imaging clinical factors. The clinical factors are integrated as the conditions of the generative modelling, which allows us to investigate how these factors influence the cardiac anatomy. We evaluate the model performance in mainly two tasks, anatomical sequence completion and sequence generation. The model achieves a high performance in anatomical sequence completion, comparable to or outperforming other state-of-the-art generative models. In terms of sequence generation, given clinical conditions, the model can generate realistic synthetic 4D sequential anatomies that share similar distributions with the real data.
Authors: Mingjie Li, Quanshi Zhang
Recently, a series of studies have tried to extract interactions between input variables modeled by a DNN and define such interactions as concepts encoded by the DNN. However, strictly speaking, there still lacks a solid guarantee whether such interactions indeed represent meaningful concepts. Therefore, in this paper, we examine the trustworthiness of interaction concepts from four perspectives. Extensive empirical studies have verified that a well-trained DNN usually encodes sparse, transferable, and discriminative concepts, which is partially aligned with human intuition.
Authors: Qihan Ren, Huiqi Deng, Yunuo Chen, Siyu Lou, Quanshi Zhang
In this paper, we focus on mean-field variational Bayesian Neural Networks (BNNs) and explore the representation capacity of such BNNs by investigating which types of concepts are less likely to be encoded by the BNN. It has been observed and studied that a relatively small set of interactive concepts usually emerge in the knowledge representation of a sufficiently-trained neural network, and such concepts can faithfully explain the network output. Based on this, our study proves that compared to standard deep neural networks (DNNs), it is less likely for BNNs to encode complex concepts. Experiments verify our theoretical proofs. Note that the tendency to encode less complex concepts does not necessarily imply weak representation power, considering that complex concepts exhibit low generalization power and high adversarial vulnerability. The code is available at https://github.com/sjtu-xai-lab/BNN-concepts.
Authors: Kangcheng Liu, Xinhu Zheng, Chaoqun Wang, Kai Tang, Ming Liu, Baoquan Chen
Contrastive learning has recently demonstrated great potential for unsupervised pre-training in 3D scene understanding tasks. However, most existing work randomly selects point features as anchors while building contrast, leading to a clear bias toward background points that often dominate in 3D scenes. Also, object awareness and foreground-to-background discrimination are neglected, making contrastive learning less effective. To tackle these issues, we propose a general foreground-aware feature contrast FAC++ framework to learn more effective point cloud representations in pre-training. FAC++ consists of two novel contrast designs to construct more effective and informative contrast pairs. The first is building positive pairs within the same foreground segment where points tend to have the same semantics. The second is that we prevent over-discrimination between 3D segments/objects and encourage grouped foreground-to-background distinctions at the segment level with adaptive feature learning in a Siamese correspondence network, which adaptively learns feature correlations within and across point cloud views effectively. Moreover, we have designed the foreground-prompted regional sampling to enhance more balanced foreground-aware learning, which is termed FAC++. Visualization with point activation maps shows that our contrast pairs capture clear correspondences among foreground regions during pre-training. Quantitative experiments also show that FAC++ achieves superior knowledge transfer and data efficiency in various downstream 3D semantic segmentation, instance segmentation as well as object detection tasks. All codes, data, and models are available at: https://github.com/KangchengLiu/FAC_Foreground_Aware_Contrast
Authors: Lu Chen, Siyu Lou, Keyan Zhang, Jin Huang, Quanshi Zhang
The Shapley value is widely regarded as a trustworthy attribution metric. However, when people use Shapley values to explain the attribution of input variables of a deep neural network (DNN), it usually requires a very high computational cost to approximate relatively accurate Shapley values in real-world applications. Therefore, we propose a novel network architecture, the HarsanyiNet, which makes inferences on the input sample and simultaneously computes the exact Shapley values of the input variables in a single forward propagation. The HarsanyiNet is designed on the theoretical foundation that the Shapley value can be reformulated as the redistribution of Harsanyi interactions encoded by the network.
Authors: Chanhyeok Park, Minhyuk Sung
Achieving tight bounding boxes of a shape while guaranteeing complete boundness is an essential task for efficient geometric operations and unsupervised semantic part detection. But previous methods fail to achieve both full coverage and tightness. Neural-network-based methods are not suitable for these goals due to the non-differentiability of the objective, while classic iterative search methods suffer from their sensitivity to the initialization. We propose a novel framework for finding a set of tight bounding boxes of a 3D shape via over-segmentation and iterative merging and refinement. Our result shows that utilizing effective search methods with appropriate objectives is the key to producing bounding boxes with both properties. We employ an existing pre-segmentation to split the shape and obtain over-segmentation. Then, we apply hierarchical merging with our novel tightness-aware merging and stopping criteria. To overcome the sensitivity to the initialization, we also define actions to refine the bounding box parameters in an Markov Decision Process (MDP) setup with a soft reward function promoting a wider exploration. Lastly, we further improve the refinement step with Monte Carlo Tree Search (MCTS) based multi-action space exploration. By thoughtful evaluation on diverse 3D shapes, we demonstrate full coverage, tightness, and an adequate number of bounding boxes of our method without requiring any training data or supervision. It thus can be applied to various downstream tasks in computer vision and graphics.
Authors: Hanze Dong, Wei Xiong, Deepanshu Goyal, Yihan Zhang, Winnie Chow, Rui Pan, Shizhe Diao, Jipeng Zhang, Kashun Shum, Tong Zhang
Generative foundation models are susceptible to implicit biases that can arise from extensive unsupervised training data. Such biases can produce suboptimal samples, skewed outcomes, and unfairness, with potentially serious consequences. Consequently, aligning these models with human ethics and preferences is an essential step toward ensuring their responsible and effective deployment in real-world applications. Prior research has primarily employed Reinforcement Learning from Human Feedback (RLHF) to address this problem, where generative models are fine-tuned with RL algorithms guided by a human-feedback-informed reward model. However, the inefficiencies and instabilities associated with RL algorithms frequently present substantial obstacles to the successful alignment, necessitating the development of a more robust and streamlined approach. To this end, we introduce a new framework, Reward rAnked FineTuning (RAFT), designed to align generative models effectively. Utilizing a reward model and a sufficient number of samples, our approach selects the high-quality samples, discarding those that exhibit undesired behavior, and subsequently enhancing the model by fine-tuning on these filtered samples. Our studies show that RAFT can effectively improve the model performance in both reward learning and other automated metrics in both large language models and diffusion models.
Authors: Pengyao Xie, Zhihong Ma, Ruiming Du, Xin Yang, Haiyan Cen
Fusion of 3D and MS imaging data has a great potential for high-throughput plant phenotyping of structural and biochemical as well as physiological traits simultaneously, which is important for decision support in agriculture and for crop breeders in selecting the best genotypes. However, lacking of 3D data integrity of various plant canopy structures and low-quality of MS images caused by the complex illumination effects make a great challenge, especially at the proximal imaging scale. Therefore, this study proposed a novel approach for adaptive data acquisition and radiometric calibration to generate high-quality 3DMPCs of plants. An efficient NBV planning method based on an UGV plant phenotyping system with a multi-sensor-equipped robotic arm was proposed to achieve adaptive data acquisition. The NeREF was employed to predict the DN values of the hemispherical reference for radiometric calibration. For NBV planning, the average total time for single plant at a joint speed of 1.55 rad/s was about 62.8 s, with an average reduction of 18.0% compared to the unplanned. The integrity of the whole-plant data was improved by an average of 23.6% compared to the fixed viewpoints alone. Compared with the ASD measurements, the RMSE of the reflectance spectra obtained from 3DMPCs at different regions of interest was 0.08 with an average decrease of 58.93% compared to the results obtained from the single-frame of MS images without 3D radiometric calibration. The 3D-calibrated plant 3DMPCs improved the predictive accuracy of PLSR for chlorophyll content, with an average increase of 0.07 in R2 and an average decrease of 21.25% in RMSE. Our approach introduced a fresh perspective on generating high-quality 3DMPCs of plants under the natural light condition, enabling more precise analysis of plant morphological and physiological parameters.
Authors: Danyal F. Bhutto, Bo Zhu, Jeremiah Z. Liu, Neha Koonjoo, Hongwei B. Li, Bruce R. Rosen, Matthew S. Rosen
Accurate image reconstruction is at the heart of diagnostics in medical imaging. Supervised deep learning-based approaches have been investigated for solving inverse problems including image reconstruction. However, these trained models encounter unseen data distributions that are widely shifted from training data during deployment. Therefore, it is essential to assess whether a given input falls within the training data distribution for diagnostic purposes. Uncertainty estimation approaches exist but focus on providing an uncertainty map to radiologists, rather than assessing the training distribution fit. In this work, we propose a method based on the local Lipschitz-based metric to distinguish out-of-distribution images from in-distribution with an area under the curve of 99.94%. Empirically, we demonstrate a very strong relationship between the local Lipschitz value and mean absolute error (MAE), supported by a high Spearman's rank correlation coefficient of 0.8475, which determines the uncertainty estimation threshold for optimal model performance. Through the identification of false positives, the local Lipschitz and MAE relationship was used to guide data augmentation and reduce model uncertainty. Our study was validated using the AUTOMAP architecture for sensor-to-image Magnetic Resonance Imaging (MRI) reconstruction. We compare our proposed approach with baseline methods: Monte-Carlo dropout and deep ensembles, and further analysis included MRI denoising and Computed Tomography (CT) sparse-to-full view reconstruction using UNET architectures. We show that our approach is applicable to various architectures and learned functions, especially in the realm of medical image reconstruction, where preserving the diagnostic accuracy of reconstructed images remains paramount.
Authors: Nazmul Karim, Umar Khalid, Mohsen Joneidi, Chen Chen, Nazanin Rahnavard
Text-to-Image (T2I) diffusion models have achieved remarkable success in synthesizing high-quality images conditioned on text prompts. Recent methods have tried to replicate the success by either training text-to-video (T2V) models on a very large number of text-video pairs or adapting T2I models on text-video pairs independently. Although the latter is computationally less expensive, it still takes a significant amount of time for per-video adaption. To address this issue, we propose SAVE, a novel spectral-shift-aware adaptation framework, in which we fine-tune the spectral shift of the parameter space instead of the parameters themselves. Specifically, we take the spectral decomposition of the pre-trained T2I weights and only update the singular values while freezing the corresponding singular vectors. In addition, we introduce a spectral shift regularizer aimed at placing tighter constraints on larger singular values compared to smaller ones. This form of regularization enables the model to grasp finer details within the video that align with the provided textual descriptions. We also offer theoretical justification for our proposed regularization technique. Since we are only dealing with spectral shifts, the proposed method reduces the adaptation time significantly (approx. 10 times) and has fewer resource constraints for training. Such attributes posit SAVE to be more suitable for real-world applications, e.g. editing undesirable content during video streaming. We validate the effectiveness of SAVE with an extensive experimental evaluation under different settings, e.g. style transfer, object replacement, privacy preservation, etc.
Authors: Matthew Lyon, Paul Armitage, Mauricio A Álvarez
Diffusion MRI (dMRI) is a widely used imaging modality, but requires long scanning times to acquire high resolution datasets. By leveraging the unique geometry present within this domain, we present a novel approach to dMRI angular super-resolution that extends upon the parametric continuous convolution (PCConv) framework. We introduce several additions to the operation including a Fourier feature mapping, global coordinates, and domain specific context. Using this framework, we build a fully parametric continuous convolution network (PCCNN) and compare against existing models. We demonstrate the PCCNN performs competitively while using significantly less parameters. Moreover, we show that this formulation generalises well to clinically relevant downstream analyses such as fixel-based analysis, and neurite orientation dispersion and density imaging.
Authors: Seokjun Choi, Seungwoo Yoon, Giljoo Nam, Seungyong Lee, Seung-Hwan Baek
Photometric stereo leverages variations in illumination conditions to reconstruct surface normals. Display photometric stereo, which employs a conventional monitor as an illumination source, has the potential to overcome limitations often encountered in bulky and difficult-to-use conventional setups. In this paper, we present differentiable display photometric stereo (DDPS), addressing an often overlooked challenge in display photometric stereo: the design of display patterns. Departing from using heuristic display patterns, DDPS learns the display patterns that yield accurate normal reconstruction for a target system in an end-to-end manner. To this end, we propose a differentiable framework that couples basis-illumination image formation with analytic photometric-stereo reconstruction. The differentiable framework facilitates the effective learning of display patterns via auto-differentiation. Also, for training supervision, we propose to use 3D printing for creating a real-world training dataset, enabling accurate reconstruction on the target real-world setup. Finally, we exploit that conventional LCD monitors emit polarized light, which allows for the optical separation of diffuse and specular reflections when combined with a polarization camera, leading to accurate normal reconstruction. Extensive evaluation of DDPS shows improved normal-reconstruction accuracy compared to heuristic patterns and demonstrates compelling properties such as robustness to pattern initialization, calibration errors, and simplifications in image formation and reconstruction.
Authors: Yu-Jen Chen, Xinrong Hu, Yiyu Shi, Tsung-Yi Ho
Magnetic resonance imaging (MRI) is commonly used for brain tumor segmentation, which is critical for patient evaluation and treatment planning. To reduce the labor and expertise required for labeling, weakly-supervised semantic segmentation (WSSS) methods with class activation mapping (CAM) have been proposed. However, existing CAM methods suffer from low resolution due to strided convolution and pooling layers, resulting in inaccurate predictions. In this study, we propose a novel CAM method, Attentive Multiple-Exit CAM (AME-CAM), that extracts activation maps from multiple resolutions to hierarchically aggregate and improve prediction accuracy. We evaluate our method on the BraTS 2021 dataset and show that it outperforms state-of-the-art methods.
Authors: Hanieh Naderi, Ivan V. Bajić
Deep learning has successfully solved a wide range of tasks in 2D vision as a dominant AI technique. Recently, deep learning on 3D point clouds is becoming increasingly popular for addressing various tasks in this field. Despite remarkable achievements, deep learning algorithms are vulnerable to adversarial attacks. These attacks are imperceptible to the human eye but can easily fool deep neural networks in the testing and deployment stage. To encourage future research, this survey summarizes the current progress on adversarial attack and defense techniques on point cloud classification.This paper first introduces the principles and characteristics of adversarial attacks and summarizes and analyzes adversarial example generation methods in recent years. Additionally, it provides an overview of defense strategies, organized into data-focused and model-focused methods. Finally, it presents several current challenges and potential future research directions in this domain.
Authors: Fei Guo, Li Zhu, YiWang Wang, Jing Sun
In the research field of few-shot learning, the main difference between image-based and video-based is the additional temporal dimension. In recent years, some works have used the Transformer to deal with frames, then get the attention feature and the enhanced prototype, and the results are competitive. However, some video frames may relate little to the action, and only using single frame-level or segment-level features may not mine enough information. We address these problems sequentially through an end-to-end method named "Task-Specific Alignment and Multiple-level Transformer Network (TSA-MLT)". The first module (TSA) aims at filtering the action-irrelevant frames for action duration alignment. Affine Transformation for frame sequence in the time dimension is used for linear sampling. The second module (MLT) focuses on the Multiple-level feature of the support prototype and query sample to mine more information for the alignment, which operates on different level features. We adopt a fusion loss according to a fusion distance that fuses the L2 sequence distance, which focuses on temporal order alignment, and the Optimal Transport distance, which focuses on measuring the gap between the appearance and semantics of the videos. Extensive experiments show our method achieves state-of-the-art results on the HMDB51 and UCF101 datasets and a competitive result on the benchmark of Kinetics and something 2-something V2 datasets. Our code is available at the URL: https://github.com/cofly2014/tsa-mlt.git
Authors: Chenyu Zhang, Zhaozheng Yin, Ruwen Qin
Efficiently monitoring the condition of civil infrastructures necessitates automating the structural condition assessment in visual inspection. This paper proposes an Attention-Enhanced Co-Interactive Fusion Network (AECIF-Net) for automatic structural condition assessment in visual bridge inspection. AECIF-Net can simultaneously parse structural elements and segment surface defects on the elements in inspection images. It integrates two task-specific relearning subnets to extract task-specific features from an overall feature embedding. A co-interactive feature fusion module further captures the spatial correlation and facilitates information sharing between tasks. Experimental results demonstrate that the proposed AECIF-Net outperforms the current state-of-the-art approaches, achieving promising performance with 92.11% mIoU for element segmentation and 87.16% mIoU for corrosion segmentation on the test set of the new benchmark dataset Steel Bridge Condition Inspection Visual (SBCIV). An ablation study verifies the merits of the designs for AECIF-Net, and a case study demonstrates its capability to automate structural condition assessment.
Authors: Yan Ma, Weicong Liang, Bohan Chen, Yiduo Hao, Bojian Hou, Xiangyu Yue, Chao Zhang, Yuhui Yuan
Motivated by the remarkable achievements of DETR-based approaches on COCO object detection and segmentation benchmarks, recent endeavors have been directed towards elevating their performance through self-supervised pre-training of Transformers while preserving a frozen backbone. Noteworthy advancements in accuracy have been documented in certain studies. Our investigation delved deeply into a representative approach, DETReg, and its performance assessment in the context of emerging models like $\mathcal{H}$-Deformable-DETR. Regrettably, DETReg proves inadequate in enhancing the performance of robust DETR-based models under full data conditions. To dissect the underlying causes, we conduct extensive experiments on COCO and PASCAL VOC probing elements such as the selection of pre-training datasets and strategies for pre-training target generation. By contrast, we employ an optimized approach named Simple Self-training which leads to marked enhancements through the combination of an improved box predictor and the Objects$365$ benchmark. The culmination of these endeavors results in a remarkable AP score of $59.3\%$ on the COCO val set, outperforming $\mathcal{H}$-Deformable-DETR + Swin-L without pre-training by $1.4\%$. Moreover, a series of synthetic pre-training datasets, generated by merging contemporary image-to-text(LLaVA) and text-to-image (SDXL) models, significantly amplifies object detection capabilities.
Authors: Fangrui Zhu, Yiming Xie, Weidi Xie, Huaizu Jiang
We have witnessed significant progress in human-object interaction (HOI) detection. The reliance on mAP (mean Average Precision) scores as a summary metric, however, does not provide sufficient insight into the nuances of model performance (e.g., why one model is better than another), which can hinder further innovation in this field. To address this issue, in this paper, we introduce a diagnosis toolbox to provide detailed quantitative break-down analysis of HOI detection models, inspired by the success of object detection diagnosis toolboxes. We first conduct holistic investigations in the pipeline of HOI detection. By defining a set of errors and the oracles to fix each of them, we can have a quantitative analysis of the significance of different errors according to the mAP improvement obtained from fixing each error. We then delve into two sub-tasks of HOI detection: human-object pair detection and interaction classification, respectively. For the first detection task, we compute the coverage of ground-truth human-object pairs as well as the noisiness level in the detection results. For the second classification task, we measure a model's performance of differentiating positive and negative detection results and also classifying the actual interactions when the human-object pairs are correctly detected. We analyze eight state-of-the-art HOI detection models and provide valuable diagnosis insights to foster future research. For instance, our diagnosis shows that state-of-the-art model RLIPv2 outperforms others mainly because it significantly improves the multi-label interaction classification accuracy. Our toolbox is applicable for different methods across different datasets and available at https://github.com/neu-vi/Diag-HOI.
Authors: Jingfan Tan, Xiaoxu Chen, Tao Wang, Kaihao Zhang, Wenhan Luo, Xiaocun Cao
By hiding the front-facing camera below the display panel, Under-Display Camera (UDC) provides users with a full-screen experience. However, due to the characteristics of the display, images taken by UDC suffer from significant quality degradation. Methods have been proposed to tackle UDC image restoration and advances have been achieved. There are still no specialized methods and datasets for restoring UDC face images, which may be the most common problem in the UDC scene. To this end, considering color filtering, brightness attenuation, and diffraction in the imaging process of UDC, we propose a two-stage network UDC Degradation Model Network named UDC-DMNet to synthesize UDC images by modeling the processes of UDC imaging. Then we use UDC-DMNet and high-quality face images from FFHQ and CelebA-Test to create UDC face training datasets FFHQ-P/T and testing datasets CelebA-Test-P/T for UDC face restoration. We propose a novel dictionary-guided transformer network named DGFormer. Introducing the facial component dictionary and the characteristics of the UDC image in the restoration makes DGFormer capable of addressing blind face restoration in UDC scenarios. Experiments show that our DGFormer and UDC-DMNet achieve state-of-the-art performance.
Authors: Zitao Wang, Qiguang Miao, Peipei Zhao, Yue Xi
The portrait matting task aims to extract an alpha matte with complete semantics and finely-detailed contours. In comparison to CNN-based approaches, transformers with self-attention module have a better capacity to capture long-range dependencies and low-frequency semantic information of a portrait. However, the recent research shows that self-attention mechanism struggles with modeling high-frequency contour information and capturing fine contour details, which can lead to bias while predicting the portrait's contours. To deal with this issue, we propose EFormer to enhance the model's attention towards both of the low-frequency semantic and high-frequency contour features. For the high-frequency contours, our research demonstrates that cross-attention module between different resolutions can guide our model to allocate attention appropriately to these contour regions. Supported on this, we can successfully extract the high-frequency detail information around the portrait's contours, which are previously ignored by self-attention. Based on cross-attention module, we further build a semantic and contour detector (SCD) to accurately capture both of the low-frequency semantic and high-frequency contour features. And we design contour-edge extraction branch and semantic extraction branch to extract refined high-frequency contour features and complete low-frequency semantic information, respectively. Finally, we fuse the two kinds of features and leverage segmentation head to generate a predicted portrait matte. Experiments on VideoMatte240K (JPEG SD Format) and Adobe Image Matting (AIM) datasets demonstrate that EFormer outperforms previous portrait matte methods.
Authors: Juyoung Yun
Activation functions are the linchpins of deep learning, profoundly influencing both the representational capacity and training dynamics of neural networks. They shape not only the nature of representations but also optimize convergence rates and enhance generalization potential. Appreciating this critical role, we present the Linear Oscillation (LoC) activation function, defined as $f(x) = x \times \sin(\alpha x + \beta)$. Distinct from conventional activation functions which primarily introduce non-linearity, LoC seamlessly blends linear trajectories with oscillatory deviations. The nomenclature "Linear Oscillation" is a nod to its unique attribute of infusing linear activations with harmonious oscillations, capturing the essence of the "Importance of Confusion". This concept of "controlled confusion" within network activations is posited to foster more robust learning, particularly in contexts that necessitate discerning subtle patterns. Our empirical studies reveal that, when integrated into diverse neural architectures, the LoC activation function consistently outperforms established counterparts like ReLU and Sigmoid. The stellar performance exhibited by the avant-garde Vision Transformer model using LoC further validates its efficacy. This study illuminates the remarkable benefits of the LoC over other prominent activation functions. It champions the notion that intermittently introducing deliberate complexity or "confusion" during training can spur more profound and nuanced learning. This accentuates the pivotal role of judiciously selected activation functions in shaping the future of neural network training.
Authors: Runsen Xu, Xiaolong Wang, Tai Wang, Yilun Chen, Jiangmiao Pang, Dahua Lin
The unprecedented advancements in Large Language Models (LLMs) have shown a profound impact on natural language processing but are yet to fully embrace the realm of 3D understanding. This paper introduces PointLLM, a preliminary effort to fill this gap, enabling LLMs to understand point clouds and offering a new avenue beyond 2D visual data. PointLLM understands colored object point clouds with human instructions and generates contextually appropriate responses, illustrating its grasp of point clouds and common sense. Specifically, it leverages a point cloud encoder with a powerful LLM to effectively fuse geometric, appearance, and linguistic information. We collect a novel dataset comprising 660K simple and 70K complex point-text instruction pairs to enable a two-stage training strategy: aligning latent spaces and subsequently instruction-tuning the unified model. To rigorously evaluate the perceptual and generalization capabilities of PointLLM, we establish two benchmarks: Generative 3D Object Classification and 3D Object Captioning, assessed through three different methods, including human evaluation, GPT-4/ChatGPT evaluation, and traditional metrics. Experimental results reveal PointLLM's superior performance over existing 2D and 3D baselines, with a notable achievement in human-evaluated object captioning tasks where it surpasses human annotators in over 50% of the samples. Codes, datasets, and benchmarks are available at https://github.com/OpenRobotLab/PointLLM .
Authors: Srikar Yellapragada, Alexandros Graikos, Prateek Prasanna, Tahsin Kurc, Joel Saltz, Dimitris Samaras
To achieve high-quality results, diffusion models must be trained on large datasets. This can be notably prohibitive for models in specialized domains, such as computational pathology. Conditioning on labeled data is known to help in data-efficient model training. Therefore, histopathology reports, which are rich in valuable clinical information, are an ideal choice as guidance for a histopathology generative model. In this paper, we introduce PathLDM, the first text-conditioned Latent Diffusion Model tailored for generating high-quality histopathology images. Leveraging the rich contextual information provided by pathology text reports, our approach fuses image and textual data to enhance the generation process. By utilizing GPT's capabilities to distill and summarize complex text reports, we establish an effective conditioning mechanism. Through strategic conditioning and necessary architectural enhancements, we achieved a SoTA FID score of 7.64 for text-to-image generation on the TCGA-BRCA dataset, significantly outperforming the closest text-conditioned competitor with FID 30.1.
Authors: Zixiang Zhou, Weiyuan Li, Baoyuan Wang
We propose MDSC(Music-Dance-Style Consistency), the first evaluation metric that assesses to what degree the dance moves and music match. Existing metrics can only evaluate the motion fidelity and diversity and the degree of rhythmic matching between music and dance. MDSC measures how stylistically correlated the generated dance motion sequences and the conditioning music sequences are. We found that directly measuring the embedding distance between motion and music is not an optimal solution. We instead tackle this through modeling it as a clustering problem. Specifically, 1) we pre-train a music encoder and a motion encoder, then 2) we learn to map and align the motion and music embedding in joint space by jointly minimizing the intra-cluster distance and maximizing the inter-cluster distance, and 3) for evaluation purposes, we encode the dance moves into embedding and measure the intra-cluster and inter-cluster distances, as well as the ratio between them. We evaluate our metric on the results of several music-conditioned motion generation methods, combined with user study, we found that our proposed metric is a robust evaluation metric in measuring the music-dance style correlation.
Authors: Noam Aigerman, Thibault Groueix
This paper proposes a fully-automatic, text-guided generative method for producing periodic, repeating, tile-able 2D art, such as the one seen on floors, mosaics, ceramics, and the work of M.C. Escher. In contrast to the standard concept of a seamless texture, i.e., square images that are seamless when tiled, our method generates non-square tilings which comprise solely of repeating copies of the same object. It achieves this by optimizing both geometry and color of a 2D mesh, in order to generate a non-square tile in the shape and appearance of the desired object, with close to no additional background details. We enable geometric optimization of tilings by our key technical contribution: an unconstrained, differentiable parameterization of the space of all possible tileable shapes for a given symmetry group. Namely, we prove that modifying the laplacian used in a 2D mesh-mapping technique - Orbifold Tutte Embedding - can achieve all possible tiling configurations for a chosen planar symmetry group. We thus consider both the mesh's tile-shape and its texture as optimizable parameters, rendering the textured mesh via a differentiable renderer. We leverage a trained image diffusion model to define a loss on the resulting image, thereby updating the mesh's parameters based on its appearance matching the text prompt. We show our method is able to produce plausible, appealing results, with non-trivial tiles, for a variety of different periodic tiling patterns.
Authors: Ruoyu Feng, Wenming Weng, Yanhui Wang, Yuhui Yuan, Jianmin Bao, Chong Luo, Zhibo Chen, Baining Guo
In this paper, we present CCEdit, a versatile generative video editing framework based on diffusion models. Our approach employs a novel trident network structure that separates structure and appearance control, ensuring precise and creative editing capabilities. Utilizing the foundational ControlNet architecture, we maintain the structural integrity of the video during editing. The incorporation of an additional appearance branch enables users to exert fine-grained control over the edited key frame. These two side branches seamlessly integrate into the main branch, which is constructed upon existing text-to-image (T2I) generation models, through learnable temporal layers. The versatility of our framework is demonstrated through a diverse range of choices in both structure representations and personalized T2I models, as well as the option to provide the edited key frame. To facilitate comprehensive evaluation, we introduce the BalanceCC benchmark dataset, comprising 100 videos and 4 target prompts for each video. Our extensive user studies compare CCEdit with eight state-of-the-art video editing methods. The outcomes demonstrate CCEdit's substantial superiority over all other methods.
Authors: Fabio Ferreira, Ivo Rapant, Frank Hutter
Many Self-Supervised Learning (SSL) methods train their models to be invariant to different "views" of an image input for which a good data augmentation pipeline is crucial. While considerable efforts were directed towards improving pre-text tasks, architectures, or robustness (e.g., Siamese networks or teacher-softmax centering), the majority of these methods remain strongly reliant on the random sampling of operations within the image augmentation pipeline, such as the random resized crop or color distortion operation. In this paper, we argue that the role of the view generation and its effect on performance has so far received insufficient attention. To address this, we propose an easy, learning-free, yet powerful Hard View Selection (HVS) strategy designed to extend the random view generation to expose the pretrained model to harder samples during SSL training. It encompasses the following iterative steps: 1) randomly sample multiple views and create pairs of two views, 2) run forward passes for each view pair on the currently trained model, 3) adversarially select the pair yielding the worst loss, and 4) run the backward pass with the selected pair. In our empirical analysis we show that under the hood, HVS increases task difficulty by controlling the Intersection over Union of views during pretraining. With only 300-epoch pretraining, HVS is able to closely rival the 800-epoch DINO baseline which remains very favorable even when factoring in the slowdown induced by the additional forwards of HVS. Additionally, HVS consistently achieves accuracy improvements on ImageNet between 0.4% and 1.9% on linear evaluation and similar improvements on transfer tasks across multiple SSL methods, such as DINO, SimSiam, iBOT, and SimCLR.
Authors: Rongzhao Zhang, Zhian Bai, Ruoying Yu, Wenrao Pang, Lingyun Wang, Lifeng Zhu, Xiaofan Zhang, Huan Zhang, Weiguo Hu
When delineating lesions from medical images, a human expert can always keep in mind the anatomical structure behind the voxels. However, although high-quality (though not perfect) anatomical information can be retrieved from computed tomography (CT) scans with modern deep learning algorithms, it is still an open problem how these automatically generated organ masks can assist in addressing challenging lesion segmentation tasks, such as the segmentation of colorectal cancer (CRC). In this paper, we develop a novel Anatomy-Guided segmentation framework to exploit the auto-generated organ masks to aid CRC segmentation from CT, namely AG-CRC. First, we obtain multi-organ segmentation (MOS) masks with existing MOS models (e.g., TotalSegmentor) and further derive a more robust organ of interest (OOI) mask that may cover most of the colon-rectum and CRC voxels. Then, we propose an anatomy-guided training patch sampling strategy by optimizing a heuristic gain function that considers both the proximity of important regions (e.g., the tumor or organs of interest) and sample diversity. Third, we design a novel self-supervised learning scheme inspired by the topology of tubular organs like the colon to boost the model performance further. Finally, we employ a masked loss scheme to guide the model to focus solely on the essential learning region. We extensively evaluate the proposed method on two CRC segmentation datasets, where substantial performance improvement (5% to 9% in Dice) is achieved over current state-of-the-art medical image segmentation models, and the ablation studies further evidence the efficacy of every proposed component.
Authors: Duy-Kien Nguyen, Martin R. Oswald, Cees G. M. Snoek
The ability to detect objects in images at varying scales has played a pivotal role in the design of modern object detectors. Despite considerable progress in removing hand-crafted components and simplifying the architecture with transformers, multi-scale feature maps and/or pyramid design remain a key factor for their empirical success. In this paper, we show that this reliance on either feature pyramids or an hierarchical backbone is unnecessary and a transformer-based detector with scale-aware attention enables the plain detector `SimPLR' whose backbone and detection head are both non-hierarchical and operate on single-scale features. The plain architecture allows SimPLR to effectively take advantages of self-supervised learning and scaling approaches with ViTs, yielding competitive performance compared to hierarchical and multi-scale counterparts. We demonstrate through our experiments that when scaling to larger ViT backbones, SimPLR indicates better performance than end-to-end segmentation models (Mask2Former) and plain-backbone detectors (ViTDet), while consistently being faster. The code will be released.
Authors: Zhikang Dong, Bin Chen, Xiulong Liu, Pawel Polak, Peng Zhang
Music recommendation for videos attracts growing interest in multi-modal research. However, existing systems focus primarily on content compatibility, often ignoring the users' preferences. Their inability to interact with users for further refinements or to provide explanations leads to a less satisfying experience. We address these issues with MuseChat, a first-of-its-kind dialogue-based recommendation system that personalizes music suggestions for videos. Our system consists of two key functionalities with associated modules: recommendation and reasoning. The recommendation module takes a video along with optional information including previous suggested music and user's preference as inputs and retrieves an appropriate music matching the context. The reasoning module, equipped with the power of Large Language Model (Vicuna-7B) and extended to multi-modal inputs, is able to provide reasonable explanation for the recommended music. To evaluate the effectiveness of MuseChat, we build a large-scale dataset, conversational music recommendation for videos, that simulates a two-turn interaction between a user and a recommender based on accurate music track information. Experiment results show that MuseChat achieves significant improvements over existing video-based music retrieval methods as well as offers strong interpretability and interactability.
Authors: Kirill Aistov, Maxim Koroteev
Based on the standard VMAF implementation we propose an implementation of VMAF using PyTorch framework. For this implementation comparisons with the standard (libvmaf) show the discrepancy $\lesssim 10^{-2}$ in VMAF units. We investigate gradients computation when using VMAF as an objective function and demonstrate that training using this function does not result in ill-behaving gradients. The implementation is then used to train a preprocessing filter. It is demonstrated that its performance is superior to the unsharp masking filter. The resulting filter is also easy for implementation and can be applied in video processing tasks for video copression improvement. This is confirmed by the results of numerical experiments.
Authors: Jiaxin Cheng, Tianjun Xiao, Tong He
We introduce a novel and efficient approach for text-based video-to-video editing that eliminates the need for resource-intensive per-video-per-model finetuning. At the core of our approach is a synthetic paired video dataset tailored for video-to-video transfer tasks. Inspired by Instruct Pix2Pix's image transfer via editing instruction, we adapt this paradigm to the video domain. Extending the Prompt-to-Prompt to videos, we efficiently generate paired samples, each with an input video and its edited counterpart. Alongside this, we introduce the Long Video Sampling Correction during sampling, ensuring consistent long videos across batches. Our method surpasses current methods like Tune-A-Video, heralding substantial progress in text-based video-to-video editing and suggesting exciting avenues for further exploration and deployment.
Authors: Xiaoting Yin, Hao Shi, Jiaan Chen, Ze Wang, Yaozu Ye, Huajian Ni, Kailun Yang, Kaiwei Wang
Human pose estimation is a fundamental and appealing task in computer vision. Traditional frame-based cameras and videos are commonly applied, yet, they become less reliable in scenarios under high dynamic range or heavy motion blur. In contrast, event cameras offer a robust solution for navigating these challenging contexts. Predominant methodologies incorporate event cameras into learning frameworks by accumulating events into event frames. However, such methods tend to marginalize the intrinsic asynchronous and high temporal resolution characteristics of events. This disregard leads to a loss in essential temporal dimension data, crucial for discerning distinct actions. To address this issue and to unlock the 3D potential of event information, we introduce two 3D event representations: the Rasterized Event Point Cloud (RasEPC) and the Decoupled Event Voxel (DEV). The RasEPC collates events within concise temporal slices at identical positions, preserving 3D attributes with statistical cues and markedly mitigating memory and computational demands. Meanwhile, the DEV representation discretizes events into voxels and projects them across three orthogonal planes, utilizing decoupled event attention to retrieve 3D cues from the 2D planes. Furthermore, we develop and release EV-3DPW, a synthetic event-based dataset crafted to facilitate training and quantitative analysis in outdoor scenes. On the public real-world DHP19 dataset, our event point cloud technique excels in real-time mobile predictions, while the decoupled event voxel method achieves the highest accuracy. Experiments on EV-3DPW demonstrate that the robustness of our proposed 3D representation methods compared to traditional RGB images and event frame techniques under the same backbones. Our code and dataset have been made publicly available at https://github.com/MasterHow/EventPointPose.
Authors: Ziyan Guo, Li Xu, Jun Liu
The rapid progress of Large Models (LMs) has recently revolutionized various fields of deep learning with remarkable grades, ranging from Natural Language Processing (NLP) to Computer Vision (CV). However, LMs are increasingly challenged and criticized by academia and industry due to their powerful performance but untrustworthy behavior, which urgently needs to be alleviated by reliable methods. Despite the abundance of literature on trustworthy LMs in NLP, a systematic survey specifically delving into the trustworthiness of LMs in CV remains absent. In order to mitigate this gap, we summarize four relevant concerns that obstruct the trustworthy usage in vision of LMs in this survey, including 1) human misuse, 2) vulnerability, 3) inherent issue and 4) interpretability. By highlighting corresponding challenge, countermeasures, and discussion in each topic, we hope this survey will facilitate readers' understanding of this field, promote alignment of LMs with human expectations and enable trustworthy LMs to serve as welfare rather than disaster for human society.
Authors: Supreeth Mysore Venkatesh, Antonio Macaluso, Marlon Nuske, Matthias Klusch, Andreas Dengel
In this study, we present Q-Seg, a novel unsupervised image segmentation method based on quantum annealing, tailored for existing quantum hardware. We formulate the pixel-wise segmentation problem, which assimilates spectral and spatial information of the image, as a graph-cut optimization task. Our method efficiently leverages the interconnected qubit topology of the D-Wave Advantage device, offering superior scalability over existing quantum approaches and outperforming state-of-the-art classical methods. Our empirical evaluations on synthetic datasets reveal that Q-Seg offers better runtime performance against the classical optimizer Gurobi. Furthermore, we evaluate our method on segmentation of Earth Observation images, an area of application where the amount of labeled data is usually very limited. In this case, Q-Seg demonstrates near-optimal results in flood mapping detection with respect to classical supervised state-of-the-art machine learning methods. Also, Q-Seg provides enhanced segmentation for forest coverage compared to existing annotated masks. Thus, Q-Seg emerges as a viable alternative for real-world applications using available quantum hardware, particularly in scenarios where the lack of labeled data and computational runtime are critical.
Authors: Martin Schilling, Christina Unterberg-Buchwald, Joachim Lotz, Martin Uecker
In recent years, a variety of deep learning networks for cardiac MRI (CMR) segmentation have been developed and analyzed. However, nearly all of them are focused on cine CMR under breathold. In this work, accuracy of deep learning methods is assessed for volumetric analysis (via segmentation) of the left ventricle in real-time free-breathing CMR at rest and under exercise stress. Data from healthy volunteers (n=15) for cine and real-time free-breathing CMR were analyzed retrospectively. Segmentations of a commercial software (comDL) and a freely available neural network (nnU-Net), were compared to a reference created via the manual correction of comDL segmentation. Segmentation of left ventricular endocardium (LV), left ventricular myocardium (MYO), and right ventricle (RV) is evaluated for both end-systolic and end-diastolic phases and analyzed with Dice's coefficient (DC). The volumetric analysis includes LV end-diastolic volume (EDV), LV end-systolic volume (ESV), and LV ejection fraction (EF). For cine CMR, nnU-Net and comDL achieve a DC above 0.95 for LV and 0.9 for MYO, and RV. For real-time CMR, the accuracy of nnU-Net exceeds that of comDL overall. For real-time CMR at rest, nnU-Net achieves a DC of 0.94 for LV, 0.89 for MYO, and 0.90 for RV; mean absolute differences between nnU-Net and reference are 2.9mL for EDV, 3.5mL for ESV and 2.6% for EF. For real-time CMR under exercise stress, nnU-Net achieves a DC of 0.92 for LV, 0.85 for MYO, and 0.83 for RV; mean absolute differences between nnU-Net and reference are 11.4mL for EDV, 2.9mL for ESV and 3.6% for EF. Deep learning methods designed or trained for cine CMR segmentation can perform well on real-time CMR. For real-time free-breathing CMR at rest, the performance of deep learning methods is comparable to inter-observer variability in cine CMR and is usable or fully automatic segmentation.
Authors: Yiwen Chen, Zilong Chen, Chi Zhang, Feng Wang, Xiaofeng Yang, Yikai Wang, Zhongang Cai, Lei Yang, Huaping Liu, Guosheng Lin
3D editing plays a crucial role in many areas such as gaming and virtual reality. Traditional 3D editing methods, which rely on representations like meshes and point clouds, often fall short in realistically depicting complex scenes. On the other hand, methods based on implicit 3D representations, like Neural Radiance Field (NeRF), render complex scenes effectively but suffer from slow processing speeds and limited control over specific scene areas. In response to these challenges, our paper presents GaussianEditor, an innovative and efficient 3D editing algorithm based on Gaussian Splatting (GS), a novel 3D representation. GaussianEditor enhances precision and control in editing through our proposed Gaussian semantic tracing, which traces the editing target throughout the training process. Additionally, we propose Hierarchical Gaussian splatting (HGS) to achieve stabilized and fine results under stochastic generative guidance from 2D diffusion models. We also develop editing strategies for efficient object removal and integration, a challenging task for existing methods. Our comprehensive experiments demonstrate GaussianEditor's superior control, efficacy, and rapid performance, marking a significant advancement in 3D editing. Project Page: https://buaacyw.github.io/gaussian-editor/
Authors: Theodor Westny, Arman Mohammadi, Daniel Jung, Erik Frisk
This paper addresses the training of Neural Ordinary Differential Equations (neural ODEs), and in particular explores the interplay between numerical integration techniques, stability regions, step size, and initialization techniques. It is shown how the choice of integration technique implicitly regularizes the learned model, and how the solver's corresponding stability region affects training and prediction performance. From this analysis, a stability-informed parameter initialization technique is introduced. The effectiveness of the initialization method is displayed across several learning benchmarks and industrial applications.
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: Konstantin Klemmer, Esther Rolf, Caleb Robinson, Lester Mackey, Marc Rußwurm
Geographic location is essential for modeling tasks in fields ranging from ecology to epidemiology to the Earth system sciences. However, extracting relevant and meaningful characteristics of a location can be challenging, often entailing expensive data fusion or data distillation from global imagery datasets. To address this challenge, we introduce Satellite Contrastive Location-Image Pretraining (SatCLIP), a global, general-purpose geographic location encoder that learns an implicit representation of locations from openly available satellite imagery. Trained location encoders provide vector embeddings summarizing the characteristics of any given location for convenient usage in diverse downstream tasks. We show that SatCLIP embeddings, pretrained on globally sampled multi-spectral Sentinel-2 satellite data, can be used in various predictive tasks that depend on location information but not necessarily satellite imagery, including temperature prediction, animal recognition in imagery, and population density estimation. Across tasks, SatCLIP embeddings consistently outperform embeddings from existing pretrained location encoders, ranging from models trained on natural images to models trained on semantic context. SatCLIP embeddings also help to improve geographic generalization. This demonstrates the potential of general-purpose location encoders and opens the door to learning meaningful representations of our planet from the vast, varied, and largely untapped modalities of geospatial data.
Authors: Daniela Massiceti, Camilla Longden, Agnieszka Słowik, Samuel Wills, Martin Grayson, Cecily Morrison
Large multi-modal models (LMMs) hold the potential to usher in a new era of automated visual assistance for people who are blind or low vision (BLV). Yet, these models have not been systematically evaluated on data captured by BLV users. We address this by empirically assessing CLIP, a widely-used LMM likely to underpin many assistive technologies. Testing 25 CLIP variants in a zero-shot classification task, we find that their accuracy is 15 percentage points lower on average for images captured by BLV users than web-crawled images. This disparity stems from CLIP's sensitivities to 1) image content (e.g. not recognizing disability objects as well as other objects); 2) image quality (e.g. not being robust to lighting variation); and 3) text content (e.g. not recognizing objects described by tactile adjectives as well as visual ones). We delve deeper with a textual analysis of three common pre-training datasets: LAION-400M, LAION-2B and DataComp-1B, showing that disability content is rarely mentioned. We then provide three examples that illustrate how the performance disparities extend to three downstream models underpinned by CLIP: OWL-ViT, CLIPSeg and DALL-E2. We find that few-shot learning with as few as 5 images can mitigate CLIP's quality-of-service disparities for BLV users in some scenarios, which we discuss alongside a set of other possible mitigations.
Authors: Haoyu Zhao, Tianyi Lu, Jiaxi Gu, Xing Zhang, Zuxuan Wu, Hang Xu, Yu-Gang Jiang
Identity-consistent video generation seeks to synthesize videos that are guided by both textual prompts and reference images of entities. Current approaches typically utilize cross-attention layers to integrate the appearance of the entity, which predominantly captures semantic attributes, resulting in compromised fidelity of entities. Moreover, these methods necessitate iterative fine-tuning for each new entity encountered, thereby limiting their applicability. To address these challenges, we introduce VideoAssembler, a novel end-to-end framework for identity-consistent video generation that can conduct inference directly when encountering new entities. VideoAssembler is adept at producing videos that are not only flexible with respect to the input reference entities but also responsive to textual conditions. Additionally, by modulating the quantity of input images for the entity, VideoAssembler enables the execution of tasks ranging from image-to-video generation to sophisticated video editing. VideoAssembler comprises two principal components: the Reference Entity Pyramid (REP) encoder and the Entity-Prompt Attention Fusion (EPAF) module. The REP encoder is designed to infuse comprehensive appearance details into the denoising stages of the stable diffusion model. Concurrently, the EPAF module is utilized to integrate text-aligned features effectively. Furthermore, to mitigate the challenge of scarce data, we present a methodology for the preprocessing of training data. Our evaluation of the VideoAssembler framework on the UCF-101, MSR-VTT, and DAVIS datasets indicates that it achieves good performances in both quantitative and qualitative analyses (346.84 in FVD and 48.01 in IS on UCF-101). Our project page is at https://gulucaptain.github.io/videoassembler/.
Authors: Duomin Wang, Bin Dai, Yu Deng, Baoyuan Wang
In this study, our goal is to create interactive avatar agents that can autonomously plan and animate nuanced facial movements realistically, from both visual and behavioral perspectives. Given high-level inputs about the environment and agent profile, our framework harnesses LLMs to produce a series of detailed text descriptions of the avatar agents' facial motions. These descriptions are then processed by our task-agnostic driving engine into motion token sequences, which are subsequently converted into continuous motion embeddings that are further consumed by our standalone neural-based renderer to generate the final photorealistic avatar animations. These streamlined processes allow our framework to adapt to a variety of non-verbal avatar interactions, both monadic and dyadic. Our extensive study, which includes experiments on both newly compiled and existing datasets featuring two types of agents -- one capable of monadic interaction with the environment, and the other designed for dyadic conversation -- validates the effectiveness and versatility of our approach. To our knowledge, we advanced a leap step by combining LLMs and neural rendering for generalized non-verbal prediction and photo-realistic rendering of avatar agents.
Authors: Subhajit Paul, Ashutosh Gupta
Clouds in optical satellite images are a major concern since their presence hinders the ability to carry accurate analysis as well as processing. Presence of clouds also affects the image tasking schedule and results in wastage of valuable storage space on ground as well as space-based systems. Due to these reasons, deriving accurate cloud masks from optical remote-sensing images is an important task. Traditional methods such as threshold-based, spatial filtering for cloud detection in satellite images suffer from lack of accuracy. In recent years, deep learning algorithms have emerged as a promising approach to solve image segmentation problems as it allows pixel-level classification and semantic-level segmentation. In this paper, we introduce a deep-learning model based on hybrid transformer architecture for effective cloud mask generation named CLiSA - Cloud segmentation via Lipschitz Stable Attention network. In this context, we propose an concept of orthogonal self-attention combined with hierarchical cross attention model, and we validate its Lipschitz stability theoretically and empirically. We design the whole setup under adversarial setting in presence of Lov\'asz-Softmax loss. We demonstrate both qualitative and quantitative outcomes for multiple satellite image datasets including Landsat-8, Sentinel-2, and Cartosat-2s. Performing comparative study we show that our model performs preferably against other state-of-the-art methods and also provides better generalization in precise cloud extraction from satellite multi-spectral (MX) images. We also showcase different ablation studies to endorse our choices corresponding to different architectural elements and objective functions.
Authors: Fukun Yin, Xin Chen, Chi Zhang, Biao Jiang, Zibo Zhao, Jiayuan Fan, Gang Yu, Taihao Li, Tao Chen
The advent of large language models, enabling flexibility through instruction-driven approaches, has revolutionized many traditional generative tasks, but large models for 3D data, particularly in comprehensively handling 3D shapes with other modalities, are still under-explored. By achieving instruction-based shape generations, versatile multimodal generative shape models can significantly benefit various fields like 3D virtual construction and network-aided design. In this work, we present ShapeGPT, a shape-included multi-modal framework to leverage strong pre-trained language models to address multiple shape-relevant tasks. Specifically, ShapeGPT employs a word-sentence-paragraph framework to discretize continuous shapes into shape words, further assembles these words for shape sentences, as well as integrates shape with instructional text for multi-modal paragraphs. To learn this shape-language model, we use a three-stage training scheme, including shape representation, multimodal alignment, and instruction-based generation, to align shape-language codebooks and learn the intricate correlations among these modalities. Extensive experiments demonstrate that ShapeGPT achieves comparable performance across shape-relevant tasks, including text-to-shape, shape-to-text, shape completion, and shape editing.
Authors: Yanqing Liu, Kai Wang, Wenqi Shao, Ping Luo, Yu Qiao, Mike Zheng Shou, Kaipeng Zhang, Yang You
Visual-language pre-training (VLP) has achieved remarkable success in multi-modal tasks, largely attributed to the availability of large-scale image-text datasets. In this work, we demonstrate that multi-modal large language models (MLLMs) can enhance visual-language representation learning by improving data quality. Our approach is simple, utilizing MLLMs to extend multiple captions for each image. To prevent the bias introduced by MLLMs' hallucinations and intrinsic caption styles, we propose "text shearing" to maintain the same length for extended captions as that of the original captions. In image-text retrieval, our method consistently obtains 5.6 ~ 35.0% and 16.8 ~ 46.1% improvement on R@1 under the fine-tuning and zero-shot settings, respectively. Notably, we obtain zero-shot results that are comparable to fine-tuning on target datasets, which encourages more exploration of the versatile use of MLLMs.