Authors: William McGough, Thomas Buddenkotte, Stephan Ursprung, Zeyu Gao, Grant Stewart, Mireia Crispin-Ortuzar
This study introduces an automated pipeline for renal cancer (RC) detection in non-contrast computed tomography (NCCT). In the development of our pipeline, we test three detections models: a shape model, a 2D-, and a 3D axial-sample model. Training (n=1348) and testing (n=64) data were gathered from open sources (KiTS23, Abdomen1k, CT-ORG) and Cambridge University Hospital (CUH). Results from cross-validation and testing revealed that the 2D axial sample model had the highest small ($\leq$40mm diameter) RC detection area under the curve (AUC) of 0.804. Our pipeline achieves 61.9\% sensitivity and 92.7\% specificity for small kidney cancers on unseen test data. Our results are much more accurate than previous attempts to automatically detect small renal cancers in NCCT, the most likely imaging modality for RC screening. This pipeline offers a promising advance that may enable screening for kidney cancers.
Authors: Anderson Augusma (M-PSI, SVH), Dominique Vaufreydaz (M-PSI), Frédérique Letué (SVH)
This paper explores privacy-compliant group-level emotion recognition ''in-the-wild'' within the EmotiW Challenge 2023. Group-level emotion recognition can be useful in many fields including social robotics, conversational agents, e-coaching and learning analytics. This research imposes itself using only global features avoiding individual ones, i.e. all features that can be used to identify or track people in videos (facial landmarks, body poses, audio diarization, etc.). The proposed multimodal model is composed of a video and an audio branches with a cross-attention between modalities. The video branch is based on a fine-tuned ViT architecture. The audio branch extracts Mel-spectrograms and feed them through CNN blocks into a transformer encoder. Our training paradigm includes a generated synthetic dataset to increase the sensitivity of our model on facial expression within the image in a data-driven way. The extensive experiments show the significance of our methodology. Our privacy-compliant proposal performs fairly on the EmotiW challenge, with 79.24% and 75.13% of accuracy respectively on validation and test set for the best models. Noticeably, our findings highlight that it is possible to reach this accuracy level with privacy-compliant features using only 5 frames uniformly distributed on the video.
Authors: Ying Wang, Yanlai Yang, Mengye Ren
The egocentric video natural language query (NLQ) task involves localizing a temporal window in an egocentric video that provides an answer to a posed query, which has wide applications in building personalized AI assistants. Prior methods for this task have focused on improvements of network architecture and leveraging pre-training for enhanced image and video features, but have struggled with capturing long-range temporal dependencies in lengthy videos, and cumbersome end-to-end training. Motivated by recent advancements in Large Language Models (LLMs) and vision language models, we introduce LifelongMemory, a novel framework that utilizes multiple pre-trained models to answer queries from extensive egocentric video content. We address the unique challenge by employing a pre-trained captioning model to create detailed narratives of the videos. These narratives are then used to prompt a frozen LLM to generate coarse-grained temporal window predictions, which are subsequently refined using a pre-trained NLQ model. Empirical results demonstrate that our method achieves competitive performance against existing supervised end-to-end learning methods, underlining the potential of integrating multiple pre-trained multimodal large language models in complex vision-language tasks. We provide a comprehensive analysis of key design decisions and hyperparameters in our pipeline, offering insights and practical guidelines.
Authors: Emre Gülsoylu, Paul Koch, Mert Yıldız, Manfred Constapel, André Peter Kelm
Deep learning object detection methods, like YOLOv5, are effective in identifying maritime vessels but often lack detailed information important for practical applications. In this paper, we addressed this problem by developing a technique that fuses Automatic Identification System (AIS) data with vessels detected in images to create datasets. This fusion enriches ship images with vessel-related data, such as type, size, speed, and direction. Our approach associates detected ships to their corresponding AIS messages by estimating distance and azimuth using a homography-based method suitable for both fixed and periodically panning cameras. This technique is useful for creating datasets for waterway traffic management, encounter detection, and surveillance. We introduce a novel dataset comprising of images taken in various weather conditions and their corresponding AIS messages. This dataset offers a stable baseline for refining vessel detection algorithms and trajectory prediction models. To assess our method's performance, we manually annotated a portion of this dataset. The results are showing an overall association accuracy of 74.76 %, with the association accuracy for fixed cameras reaching 85.06 %. This demonstrates the potential of our approach in creating datasets for vessel detection, pose estimation and auto-labelling pipelines.
Authors: Yuhang Li, Youngeun Kim, Donghyun Lee, Priyadarshini Panda
Though low-bit quantization enables efficient storage and inference of deep neural networks, it often requires the use of training data to maintain resilience against quantization errors. However, training data are frequently subject to privacy or copyright concerns. In this work, we address the challenge of Data-Scarce Quantization, where access to training data is severely limited or non-existent for quantization purposes. Conventional approaches typically rely on inverting dummy images or jointly training generative models to produce synthetic input samples. However, these methods struggle to accurately recreate complex objects in large-scale datasets like ImageNet. To overcome these limitations, we introduce StableQ, a novel method that utilizes an advanced text-to-image diffusion model to generate high-resolution, photo-realistic synthetic data. To verify the quality of the generated data, we implement two robust filtering mechanisms. These mechanisms are designed to select images that closely resemble the intrinsic characteristics of the actual training data. Furthermore, in scenarios where limited training data are available, we use these data to guide the synthetic data generation process by inverting a learnable token embedding in the text encoder. Our extensive experimental results demonstrate that StbaleQ sets a new benchmark in both zero-shot and few-shot quantization, outperforming existing methods in terms of accuracy and efficiency.
Authors: Kaiyu Song, Hanjiang Lai
Most recent works of test-time adaptation (TTA) aim to alleviate domain shift problems by re-training source classifiers in each domain. On the other hand, the emergence of the diffusion model provides another solution to TTA, which directly maps the test data from the target domain to the source domain based on a diffusion model pre-trained in the source domain. The source classifier does not need to be fine-tuned. However, 1) the semantic information loss from test data to the source domain and 2) the model shift between the source classifier and diffusion model would prevent the diffusion model from mapping the test data back to the source domain correctly. In this paper, we propose a novel guidance-based diffusion-driven adaptation (GDDA) to overcome the data shift and let the diffusion model find a better way to go back to the source. Concretely, we first propose detail and global guidance to better keep the common semantics of the test and source data. The two guidance include a contrastive loss and mean squared error to alleviate the information loss by fully exploring the diffusion model and the test data. Meanwhile, we propose a classifier-aware guidance to reduce the bias caused by the model shift, which can incorporate the source classifier's information into the generation process of the diffusion model. Extensive experiments on three image datasets with three classifier backbones demonstrate that GDDA significantly performs better than the state-of-the-art baselines. On CIFAR-10C, CIFAR-100C, and ImageNetC, GDDA achieves 11.54\%, 19.05\%, and 11.63\% average accuracy improvements, respectively. GDDA even achieves equal performance compared with methods of re-training classifiers. The code is available in the supplementary material.
Authors: Yunhao Ge, Hong-Xing Yu, Cheng Zhao, Yuliang Guo, Xinyu Huang, Liu Ren, Laurent Itti, Jiajun Wu
A major challenge in monocular 3D object detection is the limited diversity and quantity of objects in real datasets. While augmenting real scenes with virtual objects holds promise to improve both the diversity and quantity of the objects, it remains elusive due to the lack of an effective 3D object insertion method in complex real captured scenes. In this work, we study augmenting complex real indoor scenes with virtual objects for monocular 3D object detection. The main challenge is to automatically identify plausible physical properties for virtual assets (e.g., locations, appearances, sizes, etc.) in cluttered real scenes. To address this challenge, we propose a physically plausible indoor 3D object insertion approach to automatically copy virtual objects and paste them into real scenes. The resulting objects in scenes have 3D bounding boxes with plausible physical locations and appearances. In particular, our method first identifies physically feasible locations and poses for the inserted objects to prevent collisions with the existing room layout. Subsequently, it estimates spatially-varying illumination for the insertion location, enabling the immersive blending of the virtual objects into the original scene with plausible appearances and cast shadows. We show that our augmentation method significantly improves existing monocular 3D object models and achieves state-of-the-art performance. For the first time, we demonstrate that a physically plausible 3D object insertion, serving as a generative data augmentation technique, can lead to significant improvements for discriminative downstream tasks such as monocular 3D object detection. Project website: https://gyhandy.github.io/3D-Copy-Paste/
Authors: Anbo Cao, Pin-Yu Le, Zhonghui Qie, Haseeb Hassan, Yingwei Guo, Asim Zaman, Jiaxi Lu, Xueqiang Zeng, Huihui Yang, Xiaoqiang Miao, Taiyu Han, Guangtao Huang, Yan Kang, Yu Luo, Jia Guo
Dynamic susceptibility contrast magnetic resonance imaging (DSC-MRI) is widely used to evaluate acute ischemic stroke to distinguish salvageable tissue and infarct core. For this purpose, traditional methods employ deconvolution techniques, like singular value decomposition, which are known to be vulnerable to noise, potentially distorting the derived perfusion parameters. However, deep learning technology could leverage it, which can accurately estimate clinical perfusion parameters compared to traditional clinical approaches. Therefore, this study presents a perfusion parameters estimation network that considers spatial and temporal information, the Spatiotemporal Network (ST-Net), for the first time. The proposed network comprises a designed physical loss function to enhance model performance further. The results indicate that the network can accurately estimate perfusion parameters, including cerebral blood volume (CBV), cerebral blood flow (CBF), and time to maximum of the residual function (Tmax). The structural similarity index (SSIM) mean values for CBV, CBF, and Tmax parameters were 0.952, 0.943, and 0.863, respectively. The DICE score for the hypo-perfused region reached 0.859, demonstrating high consistency. The proposed model also maintains time efficiency, closely approaching the performance of commercial gold-standard software.
Authors: Jiamu Xu, Xiaoxiang Liu, Xinyuan Zhang, Yain-Whar Si, Xiaofan Li, Zheng Shi, Ke Wang, Xueyuan Gong
Learning the discriminative features of different faces is an important task in face recognition. By extracting face features in neural networks, it becomes easy to measure the similarity of different face images, which makes face recognition possible. To enhance the neural network's face feature separability, incorporating an angular margin during training is common practice. State-of-the-art loss functions CosFace and ArcFace apply fixed margins between weights of classes to enhance the inter-class separation of face features. Since the distribution of samples in the training set is imbalanced, similarities between different identities are unequal. Therefore, using an inappropriately fixed angular margin may lead to the problem that the model is difficult to converge or the face features are not discriminative enough. It is more in line with our intuition that the margins are angular adaptive, which could increase with the angles between classes growing. In this paper, we propose a new angular margin loss named X2-Softmax. X2-Softmax loss has adaptive angular margins, which provide the margin that increases with the angle between different classes growing. The angular adaptive margin ensures model flexibility and effectively improves the effect of face recognition. We have trained the neural network with X2-Softmax loss on the MS1Mv3 dataset and tested it on several evaluation benchmarks to demonstrate the effectiveness and superiority of our loss function. The experimental code and trained model are published in https://github.com/xujiamu123/X2-Softmax/tree/main.
Authors: Aël Quélennec, Enzo Tartaglione, Pavlo Mozharovskyi, Van-Tam Nguyen
In the realm of efficient on-device learning under extreme memory and computation constraints, a significant gap in successful approaches persists. Although considerable effort has been devoted to efficient inference, the main obstacle to efficient learning is the prohibitive cost of backpropagation. The resources required to compute gradients and update network parameters often exceed the limits of tightly constrained memory budgets. This paper challenges conventional wisdom and proposes a series of experiments that reveal the existence of superior sub-networks. Furthermore, we hint at the potential for substantial gains through a dynamic neuron selection strategy when fine-tuning a target task. Our efforts extend to the adaptation of a recent dynamic neuron selection strategy pioneered by Bragagnolo et al. (NEq), revealing its effectiveness in the most stringent scenarios. Our experiments demonstrate, in the average case, the superiority of a NEq-inspired approach over a random selection. This observation prompts a compelling avenue for further exploration in the area, highlighting the opportunity to design a new class of algorithms designed to facilitate parameter update selection. Our findings usher in a new era of possibilities in the field of on-device learning under extreme constraints and encourage the pursuit of innovative strategies for efficient, resource-friendly model fine-tuning.
Authors: Pratul P. Srinivasan, Stephan J. Garbin, Dor Verbin, Jonathan T. Barron, Ben Mildenhall
Existing UV mapping algorithms are designed to operate on well-behaved meshes, instead of the geometry representations produced by state-of-the-art 3D reconstruction and generation techniques. As such, applying these methods to the volume densities recovered by neural radiance fields and related techniques (or meshes triangulated from such fields) results in texture atlases that are too fragmented to be useful for tasks such as view synthesis or appearance editing. We present a UV mapping method designed to operate on geometry produced by 3D reconstruction and generation techniques. Instead of computing a mapping defined on a mesh's vertices, our method Nuvo uses a neural field to represent a continuous UV mapping, and optimizes it to be a valid and well-behaved mapping for just the set of visible points, i.e. only points that affect the scene's appearance. We show that our model is robust to the challenges posed by ill-behaved geometry, and that it produces editable UV mappings that can represent detailed appearance.
Authors: Zigeng Chen, Gongfan Fang, Xinyin Ma, Xinchao Wang
The formidable model size and demanding computational requirements of Segment Anything Model (SAM) have rendered it cumbersome for deployment on resource-constrained devices. Existing approaches for SAM compression typically involve training a new network from scratch, posing a challenging trade-off between compression costs and model performance. To address this issue, this paper introduces SlimSAM, a novel SAM compression method that achieves superior performance with remarkably low training costs. This is achieved by the efficient reuse of pre-trained SAMs through a unified pruning-distillation framework. To enhance knowledge inheritance from the original SAM, we employ an innovative alternate slimming strategy that partitions the compression process into a progressive procedure. Diverging from prior pruning techniques, we meticulously prune and distill decoupled model structures in an alternating fashion. Furthermore, a novel label-free pruning criterion is also proposed to align the pruning objective with the optimization target, thereby boosting the post-distillation after pruning. SlimSAM yields significant performance improvements while demanding over 10 times less training costs than any other existing methods. Even when compared to the original SAM-H, SlimSAM achieves approaching performance while reducing parameter counts to merely 0.9% (5.7M), MACs to 0.8% (21G), and requiring only 0.1% (10k) of the SAM training data.
Authors: Tongkun Guan, Wei Shen, Xue Yang, Xuehui Wang, Xiaokang Yang
Existing scene text detection methods typically rely on extensive real data for training. Due to the lack of annotated real images, recent works have attempted to exploit large-scale labeled synthetic data (LSD) for pre-training text detectors. However, a synth-to-real domain gap emerges, further limiting the performance of text detectors. Differently, in this work, we propose \textbf{FreeReal}, a real-domain-aligned pre-training paradigm that enables the complementary strengths of both LSD and unlabeled real data (URD). Specifically, to bridge real and synthetic worlds for pre-training, a novel glyph-based mixing mechanism (GlyphMix) is tailored for text images. GlyphMix delineates the character structures of synthetic images and embeds them as graffiti-like units onto real images. Without introducing real domain drift, GlyphMix freely yields real-world images with annotations derived from synthetic labels. Furthermore, when given free fine-grained synthetic labels, GlyphMix can effectively bridge the linguistic domain gap stemming from English-dominated LSD to URD in various languages. Without bells and whistles, FreeReal achieves average gains of 4.56\%, 3.85\%, 3.90\%, and 1.97\% in improving the performance of DBNet, PANet, PSENet, and FCENet methods, respectively, consistently outperforming previous pre-training methods by a substantial margin across four public datasets. Code will be released soon.
Authors: Gustavo Perez, Daniel Sheldon, Grant Van Horn, Subhransu Maji
Identifying the number of clusters serves as a preliminary goal for many data analysis tasks. A common approach to this problem is to vary the number of clusters in a clustering algorithm (e.g., 'k' in $k$-means) and pick the value that best explains the data. However, the count estimates can be unreliable especially when the image similarity is poor. Human feedback on the pairwise similarity can be used to improve the clustering, but existing approaches do not guarantee accurate count estimates. We propose an approach to produce estimates of the cluster counts in a large dataset given an approximate pairwise similarity. Our framework samples edges guided by the pairwise similarity, and we collect human feedback to construct a statistical estimate of the cluster count. On the technical front we have developed a nested importance sampling approach that yields (asymptotically) unbiased estimates of the cluster count with confidence intervals which can guide human effort. Compared to naive sampling, our similarity-driven sampling produces more accurate estimates of counts and tighter confidence intervals. We evaluate our method on a benchmark of six fine-grained image classification datasets achieving low error rates on the estimated number of clusters with significantly less human labeling effort compared to baselines and alternative active clustering approaches.
Authors: Yuxin Zhang, Fan Tang, Nisha Huang, Haibin Huang, Chongyang Ma, Weiming Dong, Changsheng Xu
The essence of a video lies in its dynamic motions, including character actions, object movements, and camera movements. While text-to-video generative diffusion models have recently advanced in creating diverse contents, controlling specific motions through text prompts remains a significant challenge. A primary issue is the coupling of appearance and motion, often leading to overfitting on appearance. To tackle this challenge, we introduce MotionCrafter, a novel one-shot instance-guided motion customization method. MotionCrafter employs a parallel spatial-temporal architecture that injects the reference motion into the temporal component of the base model, while the spatial module is independently adjusted for character or style control. To enhance the disentanglement of motion and appearance, we propose an innovative dual-branch motion disentanglement approach, comprising a motion disentanglement loss and an appearance prior enhancement strategy. During training, a frozen base model provides appearance normalization, effectively separating appearance from motion and thereby preserving diversity. Comprehensive quantitative and qualitative experiments, along with user preference tests, demonstrate that MotionCrafter can successfully integrate dynamic motions while preserving the coherence and quality of the base model with a wide range of appearance generation capabilities. Codes are available at https://github.com/zyxElsa/MotionCrafter.
Authors: Chen Li, Bipin Rajendran
Recent strides in low-latency spiking neural network (SNN) algorithms have drawn significant interest, particularly due to their event-driven computing nature and fast inference capability. One of the most efficient ways to construct a low-latency SNN is by converting a pre-trained, low-bit artificial neural network (ANN) into an SNN. However, this conversion process faces two main challenges: First, converting SNNs from low-bit ANNs can lead to ``occasional noise" -- the phenomenon where occasional spikes are generated in spiking neurons where they should not be -- during inference, which significantly lowers SNN accuracy. Second, although low-latency SNNs initially show fast improvements in accuracy with time steps, these accuracy growths soon plateau, resulting in their peak accuracy lagging behind both full-precision ANNs and traditional ``long-latency SNNs'' that prioritize precision over speed.
In response to these two challenges, this paper introduces a novel technique named ``noise adaptor.'' Noise adaptor can model occasional noise during training and implicitly optimize SNN accuracy, particularly at high simulation times $T$. Our research utilizes the ResNet model for a comprehensive analysis of the impact of the noise adaptor on low-latency SNNs. The results demonstrate that our method outperforms the previously reported quant-ANN-to-SNN conversion technique. We achieved an accuracy of 95.95\% within 4 time steps on CIFAR-10 using ResNet-18, and an accuracy of 74.37\% within 64 time steps on ImageNet using ResNet-50. Remarkably, these results were obtained without resorting to any noise correction methods during SNN inference, such as negative spikes or two-stage SNN simulations. Our approach significantly boosts the peak accuracy of low-latency SNNs, bringing them on par with the accuracy of full-precision ANNs. Code will be open source.
Authors: Mohammad Reza Taesiri, Tianjun Feng, Cor-Paul Bezemer, Anh Nguyen
Large multimodal models (LMMs) have evolved from large language models (LLMs) to integrate multiple input modalities, such as visual inputs. This integration augments the capacity of LLMs for tasks requiring visual comprehension and reasoning. However, the extent and limitations of their enhanced abilities are not fully understood, especially when it comes to real-world tasks. To address this gap, we introduce GlitchBench, a novel benchmark derived from video game quality assurance tasks, to test and evaluate the reasoning capabilities of LMMs. Our benchmark is curated from a variety of unusual and glitched scenarios from video games and aims to challenge both the visual and linguistic reasoning powers of LMMs in detecting and interpreting out-of-the-ordinary events. We evaluate multiple state-of-the-art LMMs, and we show that GlitchBench presents a new challenge for these models. Code and data are available at: https://glitchbench.github.io/
Authors: Jionghao Wang, Yuan Liu, Zhiyang Dou, Zhengming Yu, Yongqing Liang, Xin Li, Wenping Wang, Rong Xie, Li Song
In this paper, we introduced a novel text-to-avatar generation method that separately generates the human body and the clothes and allows high-quality animation on the generated avatar. While recent advancements in text-to-avatar generation have yielded diverse human avatars from text prompts, these methods typically combine all elements-clothes, hair, and body-into a single 3D representation. Such an entangled approach poses challenges for downstream tasks like editing or animation. To overcome these limitations, we propose a novel disentangled 3D avatar representation named Sequentially Offset-SMPL (SO-SMPL), building upon the SMPL model. SO-SMPL represents the human body and clothes with two separate meshes, but associates them with offsets to ensure the physical alignment between the body and the clothes. Then, we design an Score Distillation Sampling(SDS)-based distillation framework to generate the proposed SO-SMPL representation from text prompts. In comparison with existing text-to-avatar methods, our approach not only achieves higher exture and geometry quality and better semantic alignment with text prompts, but also significantly improves the visual quality of character animation, virtual try-on, and avatar editing. Our project page is at https://shanemankiw.github.io/SO-SMPL/.
Authors: Jalees Nehvi, Berna Kabadayi, Julien Valentin, Justus Thies
We propose 360{\deg} Volumetric Portrait (3VP) Avatar, a novel method for reconstructing 360{\deg} photo-realistic portrait avatars of human subjects solely based on monocular video inputs. State-of-the-art monocular avatar reconstruction methods rely on stable facial performance capturing. However, the common usage of 3DMM-based facial tracking has its limits; side-views can hardly be captured and it fails, especially, for back-views, as required inputs like facial landmarks or human parsing masks are missing. This results in incomplete avatar reconstructions that only cover the frontal hemisphere. In contrast to this, we propose a template-based tracking of the torso, head and facial expressions which allows us to cover the appearance of a human subject from all sides. Thus, given a sequence of a subject that is rotating in front of a single camera, we train a neural volumetric representation based on neural radiance fields. A key challenge to construct this representation is the modeling of appearance changes, especially, in the mouth region (i.e., lips and teeth). We, therefore, propose a deformation-field-based blend basis which allows us to interpolate between different appearance states. We evaluate our approach on captured real-world data and compare against state-of-the-art monocular reconstruction methods. In contrast to those, our method is the first monocular technique that reconstructs an entire 360{\deg} avatar.
Authors: Ribana Roscher, Marc Rußwurm, Caroline Gevaert, Michael Kampffmeyer, Jefersson A. dos Santos, Maria Vakalopoulou, Ronny Hänsch, Stine Hansen, Keiller Nogueira, Jonathan Prexl, Devis Tuia
Recent developments and research in modern machine learning have led to substantial improvements in the geospatial field. Although numerous deep learning models have been proposed, the majority of them have been developed on benchmark datasets that lack strong real-world relevance. Furthermore, the performance of many methods has already saturated on these datasets. We argue that shifting the focus towards a complementary data-centric perspective is necessary to achieve further improvements in accuracy, generalization ability, and real impact in end-user applications. This work presents a definition and precise categorization of automated data-centric learning approaches for geospatial data. It highlights the complementary role of data-centric learning with respect to model-centric in the larger machine learning deployment cycle. We review papers across the entire geospatial field and categorize them into different groups. A set of representative experiments shows concrete implementation examples. These examples provide concrete steps to act on geospatial data with data-centric machine learning approaches.
Authors: Florian Barthel, Anna Hilsmann, Peter Eisert
Current 3D GAN inversion methods for human heads typically use only one single frontal image to reconstruct the whole 3D head model. This leaves out meaningful information when multi-view data or dynamic videos are available. Our method builds on existing state-of-the-art 3D GAN inversion techniques to allow for consistent and simultaneous inversion of multiple views of the same subject. We employ a multi-latent extension to handle inconsistencies present in dynamic face videos to re-synthesize consistent 3D representations from the sequence. As our method uses additional information about the target subject, we observe significant enhancements in both geometric accuracy and image quality, particularly when rendering from wide viewing angles. Moreover, we demonstrate the editability of our inverted 3D renderings, which distinguishes them from NeRF-based scene reconstructions.
Authors: Sulaiman Vesal, Indrani Bhattacharya, Hassan Jahanandish, Xinran Li, Zachary Kornberg, Steve Ran Zhou, Elijah Richard Sommer, Moon Hyung Choi, Richard E. Fan, Geoffrey A. Sonn, Mirabela Rusu
Interpreting traditional B-mode ultrasound images can be challenging due to image artifacts (e.g., shadowing, speckle), leading to low sensitivity and limited diagnostic accuracy. While Magnetic Resonance Imaging (MRI) has been proposed as a solution, it is expensive and not widely available. Furthermore, most biopsies are guided by Transrectal Ultrasound (TRUS) alone and can miss up to 52% cancers, highlighting the need for improved targeting. To address this issue, we propose ProsDectNet, a multi-task deep learning approach that localizes prostate cancer on B-mode ultrasound. Our model is pre-trained using radiologist-labeled data and fine-tuned using biopsy-confirmed labels. ProsDectNet includes a lesion detection and patch classification head, with uncertainty minimization using entropy to improve model performance and reduce false positive predictions. We trained and validated ProsDectNet using a cohort of 289 patients who underwent MRI-TRUS fusion targeted biopsy. We then tested our approach on a group of 41 patients and found that ProsDectNet outperformed the average expert clinician in detecting prostate cancer on B-mode ultrasound images, achieving a patient-level ROC-AUC of 82%, a sensitivity of 74%, and a specificity of 67%. Our results demonstrate that ProsDectNet has the potential to be used as a computer-aided diagnosis system to improve targeted biopsy and treatment planning.
Authors: Zahra Nabizadeh Shahre-Babak, Nader Karimi, Krishna Rapaka, Tarek Amara, Shadrokh Samavi, Shahram Shirani
Video streaming often requires transcoding content into different resolutions and bitrates to match the recipient's internet speed and screen capabilities. Video encoders like x264 offer various presets, each with different tradeoffs between transcoding time and rate-distortion performance. Choosing the best preset for video transcoding is difficult, especially for live streaming, as trying all the presets and choosing the best one is not feasible. One solution is to predict each preset's transcoding time and select the preset that ensures the highest quality while adhering to live streaming time constraints. Prediction of video transcoding time is also critical in minimizing streaming delays, deploying resource management algorithms, and load balancing. We propose a learning-based framework for predicting the transcoding time of videos across various presets. Our predictor's features for video transcoding time prediction are derived directly from the ingested stream, primarily from the header or metadata. As a result, only minimal additional delay is incurred for feature extraction, rendering our approach ideal for live-streaming applications. We evaluated our learning-based transcoding time prediction using a dataset of videos. The results demonstrate that our framework can accurately predict the transcoding time for different presets, with a mean absolute percentage error (MAPE) of nearly 5.0%. Leveraging these predictions, we then select the most suitable transcoding preset for live video streaming. Utilizing our transcoding time prediction-based preset selection improved Peak Signal-to-Noise Ratio (PSNR) of up to 5 dB.
Authors: Diego Bonilla
In the domain of vision-language integration, generating detailed image captions poses a significant challenge due to the lack of a curated and rich dataset. This study introduces PixLore, a novel method that leverages Querying Transformers through the fine-tuning of the BLIP-2 model using the LoRa method on a standard commercial GPU. Our approach, which involves training on a carefully assembled dataset from state-of-the-art Computer Vision models combined and augmented by ChatGPT, addresses the question of whether intricate image understanding can be achieved with an ensemble of smaller-scale models. Comparative evaluations against major models such as GPT-4 and Google Bard demonstrate that PixLore-2.7B, despite having considerably fewer parameters, is rated higher than the existing State-of-the-Art models in over half of the assessments. This research not only presents a groundbreaking approach but also highlights the importance of well-curated datasets in enhancing the performance of smaller models.
Authors: Marios S. Pattichis, Venkatesh Jatla, Alvaro E. Ullao Cerna
The paper provides a survey of the development of machine-learning techniques for video analysis. The survey provides a summary of the most popular deep learning methods used for human activity recognition. We discuss how popular architectures perform on standard datasets and highlight the differences from real-life datasets dominated by multiple activities performed by multiple participants over long periods. For real-life datasets, we describe the use of low-parameter models (with 200X or 1,000X fewer parameters) that are trained to detect a single activity after the relevant objects have been successfully detected. Our survey then turns to a summary of machine learning methods that are specifically developed for working with a small number of labeled video samples. Our goal here is to describe modern techniques that are specifically designed so as to minimize the amount of ground truth that is needed for training and testing video analysis systems. We provide summaries of the development of self-supervised learning, semi-supervised learning, active learning, and zero-shot learning for applications in video analysis. For each method, we provide representative examples.
Authors: Felix A. Wichmann, Simon Kornblith, Robert Geirhos
Neither the hype exemplified in some exaggerated claims about deep neural networks (DNNs), nor the gloom expressed by Bowers et al. do DNNs as models in vision science justice: DNNs rapidly evolve, and today's limitations are often tomorrow's successes. In addition, providing explanations as well as prediction and image-computability are model desiderata; one should not be favoured at the expense of the other.
Authors: Catherine Bouchard, Vincent Boulanger, Flavie Lavoie-Cardinal, Christian Gagné
Measurements of different overlapping components require robust unmixing algorithms to convert the raw multi-dimensional measurements to useful unmixed images. Such algorithms perform reliable separation of the components when the raw signal is fully resolved and contains enough information to fit curves on the raw distributions. In experimental physics, measurements are often noisy, undersampled, or unresolved spatially or spectrally. We propose a novel method where bandpass filters are applied to the latent space of a multi-dimensional convolutional neural network to separate the overlapping signal components and extract each of their relative contributions. Simultaneously processing all dimensions with multi-dimensional convolution kernels empowers the network to combine the information from adjacent pixels and time- or spectral-bins, facilitating component separation in instances where individual pixels lack well-resolved information. We demonstrate the applicability of the method to real experimental physics problems using fluorescence lifetime microscopy and mode decomposition in optical fibers as test cases. The successful application of our approach to these two distinct experimental cases, characterized by different measured distributions, highlights the versatility of our approach in addressing a wide array of imaging tasks.
Authors: Yusuf Dalva, Pinar Yanardag
Generative models have been very popular in the recent years for their image generation capabilities. GAN-based models are highly regarded for their disentangled latent space, which is a key feature contributing to their success in controlled image editing. On the other hand, diffusion models have emerged as powerful tools for generating high-quality images. However, the latent space of diffusion models is not as thoroughly explored or understood. Existing methods that aim to explore the latent space of diffusion models usually relies on text prompts to pinpoint specific semantics. However, this approach may be restrictive in areas such as art, fashion, or specialized fields like medicine, where suitable text prompts might not be available or easy to conceive thus limiting the scope of existing work. In this paper, we propose an unsupervised method to discover latent semantics in text-to-image diffusion models without relying on text prompts. Our method takes a small set of unlabeled images from specific domains, such as faces or cats, and a pre-trained diffusion model, and discovers diverse semantics in unsupervised fashion using a contrastive learning objective. Moreover, the learned directions can be applied simultaneously, either within the same domain (such as various types of facial edits) or across different domains (such as applying cat and face edits within the same image) without interfering with each other. Our extensive experiments show that our method achieves highly disentangled edits, outperforming existing approaches in both diffusion-based and GAN-based latent space editing methods.
Authors: Reza Azad, Moein Heidary, Kadir Yilmaz, Michael Hüttemann, Sanaz Karimijafarbigloo, Yuli Wu, Anke Schmeink, Dorit Merhof
Semantic image segmentation, the process of classifying each pixel in an image into a particular class, plays an important role in many visual understanding systems. As the predominant criterion for evaluating the performance of statistical models, loss functions are crucial for shaping the development of deep learning-based segmentation algorithms and improving their overall performance. To aid researchers in identifying the optimal loss function for their particular application, this survey provides a comprehensive and unified review of $25$ loss functions utilized in image segmentation. We provide a novel taxonomy and thorough review of how these loss functions are customized and leveraged in image segmentation, with a systematic categorization emphasizing their significant features and applications. Furthermore, to evaluate the efficacy of these methods in real-world scenarios, we propose unbiased evaluations of some distinct and renowned loss functions on established medical and natural image datasets. We conclude this review by identifying current challenges and unveiling future research opportunities. Finally, we have compiled the reviewed studies that have open-source implementations on our GitHub page.
Authors: Md Shazid Islam, Sayak Nag, Arindam Dutta, Miraj Ahmed, Fahim Faisal Niloy, Amit K.Roy-Chowdhury
Data privacy, storage, and distribution shifts are major bottlenecks in medical image analysis. Data cannot be shared across patients, physicians, and facilities due to privacy concerns, usually requiring each patient's data to be analyzed in a discreet setting at a near real-time pace. However, one would like to take advantage of the accumulated knowledge across healthcare facilities as the computational systems analyze data of more and more patients while incorporating feedback provided by physicians to improve accuracy. Motivated by these, we propose a method for medical image segmentation that adapts to each incoming data batch (online adaptation), incorporates physician feedback through active learning, and assimilates knowledge across facilities in a federated setup. Combining an online adaptation scheme at test time with an efficient sampling strategy with budgeted annotation helps bridge the gap between the source and the incoming stream of target domain data. A federated setup allows collaborative aggregation of knowledge across distinct distributed models without needing to share the data across different models. This facilitates the improvement of performance over time by accumulating knowledge across users. Towards achieving these goals, we propose a computationally amicable, privacy-preserving image segmentation technique \textbf{DrFRODA} that uses federated learning to adapt the model in an online manner with feedback from doctors in the loop. Our experiments on publicly available datasets show that the proposed distributed active learning-based online adaptation method outperforms unsupervised online adaptation methods and shows competitive results with offline active learning-based adaptation methods.
Authors: Ruihan Yang, Hannes Gamper, Sebastian Braun
We introduce a multi-modal diffusion model tailored for the bi-directional conditional generation of video and audio. Recognizing the importance of accurate alignment between video and audio events in multi-modal generation tasks, we propose a joint contrastive training loss to enhance the synchronization between visual and auditory occurrences. Our research methodology involves conducting comprehensive experiments on multiple datasets to thoroughly evaluate the efficacy of our proposed model. The assessment of generation quality and alignment performance is carried out from various angles, encompassing both objective and subjective metrics. Our findings demonstrate that the proposed model outperforms the baseline, substantiating its effectiveness and efficiency. Notably, the incorporation of the contrastive loss results in improvements in audio-visual alignment, particularly in the high-correlation video-to-audio generation task. These results indicate the potential of our proposed model as a robust solution for improving the quality and alignment of multi-modal generation, thereby contributing to the advancement of video and audio conditional generation systems.
Authors: Vasil Kolev, Todor Cooklev, Fritz Keinert
Para-Hermitian polynomial matrices obtained by matrix spectral factorization lead to functions useful in control theory systems, basis functions in numerical methods or multiscaling functions used in signal processing. We introduce a fast algorithm for matrix spectral factorization based on Bauer$'$s method. We convert Bauer$'$ method into a nonlinear matrix equation (NME). The NME is solved by two different numerical algorithms (Fixed Point Iteration and Newton$'$s Method) which produce approximate scalar or matrix factors, as well as a symbolic algorithm which produces exact factors in closed form for some low-order scalar or matrix polynomial matrices, respectively. Convergence rates of the two numerical algorithms are investigated for a number of singular and nonsingular scalar and matrix polynomials taken from different areas. In particular, one of the singular examples leads to new orthogonal multiscaling and multiwavelet filters. Since the NME can also be solved as a Generalized Discrete Time Algebraic Riccati Equation (GDARE), numerical results using built-in routines in Maple 17.0 and 6 Matlab versions are presented.
Authors: Xingjian Diao, Ming Cheng, Wayner Barrios, SouYoung Jin
Talking face generation has gained immense popularity in the computer vision community, with various applications including AR/VR, teleconferencing, digital assistants, and avatars. Traditional methods are mainly audio-driven ones which have to deal with the inevitable resource-intensive nature of audio storage and processing. To address such a challenge, we propose FT2TF - First-Person Statement Text-To-Talking Face Generation, a novel one-stage end-to-end pipeline for talking face generation driven by first-person statement text. Moreover, FT2TF implements accurate manipulation of the facial expressions by altering the corresponding input text. Different from previous work, our model only leverages visual and textual information without any other sources (e.g. audio/landmark/pose) during inference. Extensive experiments are conducted on LRS2 and LRS3 datasets, and results on multi-dimensional evaluation metrics are reported. Both quantitative and qualitative results showcase that FT2TF outperforms existing relevant methods and reaches the state-of-the-art. This achievement highlights our model capability to bridge first-person statements and dynamic face generation, providing insightful guidance for future work.
Authors: Yuewei Yang, Xiaoliang Dai, Jialiang Wang, Peizhao Zhang, Hongbo Zhang
Latent Diffusion Models (LDMs) capture the dynamic evolution of latent variables over time, blending patterns and multimodality in a generative system. Despite the proficiency of LDM in various applications, such as text-to-image generation, facilitated by robust text encoders and a variational autoencoder, the critical need to deploy large generative models on edge devices compels a search for more compact yet effective alternatives. Post Training Quantization (PTQ), a method to compress the operational size of deep learning models, encounters challenges when applied to LDM due to temporal and structural complexities. This study proposes a quantization strategy that efficiently quantize LDMs, leveraging Signal-to-Quantization-Noise Ratio (SQNR) as a pivotal metric for evaluation. By treating the quantization discrepancy as relative noise and identifying sensitive part(s) of a model, we propose an efficient quantization approach encompassing both global and local strategies. The global quantization process mitigates relative quantization noise by initiating higher-precision quantization on sensitive blocks, while local treatments address specific challenges in quantization-sensitive and time-sensitive modules. The outcomes of our experiments reveal that the implementation of both global and local treatments yields a highly efficient and effective Post Training Quantization (PTQ) of LDMs.
Authors: Yin Chen, Jia Li, Shiguang Shan, Meng Wang, Richang Hong
Dynamic facial expression recognition (DFER) in the wild is still hindered by data limitations, e.g., insufficient quantity and diversity of pose, occlusion and illumination, as well as the inherent ambiguity of facial expressions. In contrast, static facial expression recognition (SFER) currently shows much higher performance and can benefit from more abundant high-quality training data. Moreover, the appearance features and dynamic dependencies of DFER remain largely unexplored. To tackle these challenges, we introduce a novel Static-to-Dynamic model (S2D) that leverages existing SFER knowledge and dynamic information implicitly encoded in extracted facial landmark-aware features, thereby significantly improving DFER performance. Firstly, we build and train an image model for SFER, which incorporates a standard Vision Transformer (ViT) and Multi-View Complementary Prompters (MCPs) only. Then, we obtain our video model (i.e., S2D), for DFER, by inserting Temporal-Modeling Adapters (TMAs) into the image model. MCPs enhance facial expression features with landmark-aware features inferred by an off-the-shelf facial landmark detector. And the TMAs capture and model the relationships of dynamic changes in facial expressions, effectively extending the pre-trained image model for videos. Notably, MCPs and TMAs only increase a fraction of trainable parameters (less than +10\%) to the original image model. Moreover, we present a novel Emotion-Anchors (i.e., reference samples for each emotion category) based Self-Distillation Loss to reduce the detrimental influence of ambiguous emotion labels, further enhancing our S2D. Experiments conducted on popular SFER and DFER datasets show that we achieve the state of the art.
Authors: Qian Qiao, Yu Xie, Ziyin Zeng, Fanzhang Li
Few-shot image classification aims to classify images from unseen novel classes with few samples. Recent works demonstrate that deep local descriptors exhibit enhanced representational capabilities compared to image-level features. However, most existing methods solely rely on either employing all local descriptors or directly utilizing partial descriptors, potentially resulting in the loss of crucial information. Moreover, these methods primarily emphasize the selection of query descriptors while overlooking support descriptors. In this paper, we propose a novel Task-Aware Adaptive Local Descriptors Selection Network (TALDS-Net), which exhibits the capacity for adaptive selection of task-aware support descriptors and query descriptors. Specifically, we compare the similarity of each local support descriptor with other local support descriptors to obtain the optimal support descriptor subset and then compare the query descriptors with the optimal support subset to obtain discriminative query descriptors. Extensive experiments demonstrate that our TALDS-Net outperforms state-of-the-art methods on both general and fine-grained datasets.
Authors: Gusti Ahmad Fanshuri Alfarisy, Owais Ahmed Malik, Ong Wee Hong
Open-World Recognition (OWR) is an emerging field that makes a machine learning model competent in rejecting the unknowns, managing them, and incrementally adding novel samples to the base knowledge. However, this broad objective is not practical for an agent that works on a specific task. Not all rejected samples will be used for learning continually in the future. Some novel images in the open environment may not belong to the domain of interest. Hence, identifying the unknown in the domain of interest is essential for a machine learning model to learn merely the important samples. In this study, we propose an evaluation protocol for estimating a model's capability in separating unknown in-domain (ID) and unknown out-of-domain (OOD). We evaluated using three approaches with an unknown domain and demonstrated the possibility of identifying the domain of interest using the pre-trained parameters through traditional transfer learning, Automated Machine Learning (AutoML), and Nearest Class Mean (NCM) classifier with First Integer Neighbor Clustering Hierarchy (FINCH). We experimented with five different domains: garbage, food, dogs, plants, and birds. The results show that all approaches can be used as an initial baseline yielding a good accuracy. In addition, a Balanced Accuracy (BACCU) score from a pre-trained model indicates a tendency to excel in one or more domains of interest. We observed that MobileNetV3 yielded the highest BACCU score for the garbage domain and surpassed complex models such as the transformer network. Meanwhile, our results also suggest that a strong representation in the pre-trained model is important for identifying unknown classes in the same domain. This study could open the bridge toward open-world recognition in domain-specific tasks where the relevancy of the unknown classes is vital.
Authors: Yifan Chen, Zhiyu Pan, Zhicheng Zhong, Wenxuan Guo, Jianjiang Feng, Jie Zhou
In this paper, we present a novel registration framework, HumanReg, that learns a non-rigid transformation between two human point clouds end-to-end. We introduce body prior into the registration process to efficiently handle this type of point cloud. Unlike most exsisting supervised registration techniques that require expensive point-wise flow annotations, HumanReg can be trained in a self-supervised manner benefiting from a set of novel loss functions. To make our model better converge on real-world data, we also propose a pretraining strategy, and a synthetic dataset (HumanSyn4D) consists of dynamic, sparse human point clouds and their auto-generated ground truth annotations. Our experiments shows that HumanReg achieves state-of-the-art performance on CAPE-512 dataset and gains a qualitative result on another more challenging real-world dataset. Furthermore, our ablation studies demonstrate the effectiveness of our synthetic dataset and novel loss functions. Our code and synthetic dataset is available at https://github.com/chenyifanthu/HumanReg.
Authors: Atoosa Chegini, Soheil Feizi
Deep learning models can encounter unexpected failures, especially when dealing with challenging sub-populations. One common reason for these failures is the occurrence of objects in backgrounds that are rarely seen during training. To gain a better understanding of these failure modes, human-interpretable descriptions are crucial for further analysis and improvement which is expensive. In this study, we propose an end-to-end framework that utilizes the capabilities of large language models (ChatGPT) and vision-language deep models (CLIP) to generate text descriptions of failure modes associated with spurious correlations (e.g. rarely seen backgrounds) without human-in-the-loop intervention. These descriptions can be used to generate synthetic data using generative models, such as diffusion models. The model can now use this generated data to learn from its weaknesses and enhance its performance on backgrounds that are uncommon for each class of data. Our approach serves as a broad solution, promising progress in comprehending model failure modes and strengthening deep learning models across a wide range of failure scenarios (e.g. bacckgrounds, colors) automatically in a few-shot manner. Our experiments have shown remarkable \textbf{improvements in accuracy ($\sim \textbf{21%}$)} on hard sub-populations (particularly for wrong background association) across $40$ different models, such as ResNets, EfficientNets, DenseNets, Vision Transformer (ViT), SwAVs, MoCos, DINOs, and CLIPs on various datasets such as ImageNet-1000, CIFAR-10, and CIFAR-100.
Authors: Zhiling Zheng, Zhiguo He, Omar Khattab, Nakul Rampal, Matei A. Zaharia, Christian Borgs, Jennifer T. Chayes, Omar M. Yaghi
The integration of artificial intelligence into scientific research has reached a new pinnacle with GPT-4V, a large language model featuring enhanced vision capabilities, accessible through ChatGPT or an API. This study demonstrates the remarkable ability of GPT-4V to navigate and obtain complex data for metal-organic frameworks, especially from graphical sources. Our approach involved an automated process of converting 346 scholarly articles into 6240 images, which represents a benchmark dataset in this task, followed by deploying GPT-4V to categorize and analyze these images using natural language prompts. This methodology enabled GPT-4V to accurately identify and interpret key plots integral to MOF characterization, such as nitrogen isotherms, PXRD patterns, and TGA curves, among others, with accuracy and recall above 93%. The model's proficiency in extracting critical information from these plots not only underscores its capability in data mining but also highlights its potential in aiding the creation of comprehensive digital databases for reticular chemistry. In addition, the extracted nitrogen isotherm data from the selected literature allowed for a comparison between theoretical and experimental porosity values for over 200 compounds, highlighting certain discrepancies and underscoring the importance of integrating computational and experimental data. This work highlights the potential of AI in accelerating scientific discovery and innovation, bridging the gap between computational tools and experimental research, and paving the way for more efficient, inclusive, and comprehensive scientific inquiry.
Authors: Zijian Chen, Wei Sun, Haoning Wu, Zicheng Zhang, Jun Jia, Xiongkuo Min, Guangtao Zhai, Wenjun Zhang
The proliferation of Artificial Intelligence-Generated Images (AGIs) has greatly expanded the Image Naturalness Assessment (INA) problem. Different from early definitions that mainly focus on tone-mapped images with limited distortions (e.g., exposure, contrast, and color reproduction), INA on AI-generated images is especially challenging as it has more diverse contents and could be affected by factors from multiple perspectives, including low-level technical distortions and high-level rationality distortions. In this paper, we take the first step to benchmark and assess the visual naturalness of AI-generated images. First, we construct the AI-Generated Image Naturalness (AGIN) database by conducting a large-scale subjective study to collect human opinions on the overall naturalness as well as perceptions from technical and rationality perspectives. AGIN verifies that naturalness is universally and disparately affected by both technical and rationality distortions. Second, we propose the Joint Objective Image Naturalness evaluaTor (JOINT), to automatically learn the naturalness of AGIs that aligns human ratings. Specifically, JOINT imitates human reasoning in naturalness evaluation by jointly learning both technical and rationality perspectives. Experimental results show our proposed JOINT significantly surpasses baselines for providing more subjectively consistent results on naturalness assessment. Our database and code will be released in https://github.com/zijianchen98/AGIN.
Authors: Yuming Qiao, Fanyi Wang, Jingwen Su, Yanhao Zhang, Yunjie Yu, Siyu Wu, Guo-Jun Qi
Image editing approaches with diffusion models have been rapidly developed, yet their applicability are subject to requirements such as specific editing types (e.g., foreground or background object editing, style transfer), multiple conditions (e.g., mask, sketch, caption), and time consuming fine-tuning of diffusion models. For alleviating these limitations and realizing efficient real image editing, we propose a novel editing technique that only requires an input image and target text for various editing types including non-rigid edits without fine-tuning diffusion model. Our method contains three novelties:(I) Target-text Inversion Schedule (TTIS) is designed to fine-tune the input target text embedding to achieve fast image reconstruction without image caption and acceleration of convergence.(II) Progressive Transition Scheme applies progressive linear interpolation between target text embedding and its fine-tuned version to generate transition embedding for maintaining non-rigid editing capability.(III) Balanced Attention Module (BAM) balances the tradeoff between textual description and image semantics.By the means of combining self-attention map from reconstruction process and cross-attention map from transition process, the guidance of target text embeddings in diffusion process is optimized.In order to demonstrate editing capability, effectiveness and efficiency of the proposed BARET, we have conducted extensive qualitative and quantitative experiments. Moreover, results derived from user study and ablation study further prove the superiority over other methods.
Authors: Renao Yan, Qiehe Sun, Cheng Jin, Yiqing Liu, Yonghong He, Tian Guan, Hao Chen
In computational pathology, whole slide image (WSI) classification presents a formidable challenge due to its gigapixel resolution and limited fine-grained annotations. Multiple instance learning (MIL) offers a weakly supervised solution, yet refining instance-level information from bag-level labels remains complex. While most of the conventional MIL methods use attention scores to estimate instance importance scores (IIS) which contribute to the prediction of the slide labels, these often lead to skewed attention distributions and inaccuracies in identifying crucial instances. To address these issues, we propose a new approach inspired by cooperative game theory: employing Shapley values to assess each instance's contribution, thereby improving IIS estimation. The computation of the Shapley value is then accelerated using attention, meanwhile retaining the enhanced instance identification and prioritization. We further introduce a framework for the progressive assignment of pseudo bags based on estimated IIS, encouraging more balanced attention distributions in MIL models. Our extensive experiments on CAMELYON-16, BRACS, and TCGA-LUNG datasets show our method's superiority over existing state-of-the-art approaches, offering enhanced interpretability and class-wise insights. We will release the code upon acceptance.
Authors: Shiji Zhao, Xizhe Wang, Xingxing Wei
Adversarial Training (AT) has been widely proved to be an effective method to improve the adversarial robustness against adversarial examples for Deep Neural Networks (DNNs). As a variant of AT, Adversarial Robustness Distillation (ARD) has demonstrated its superior performance in improving the robustness of small student models with the guidance of large teacher models. However, both AT and ARD encounter the robust fairness problem: these models exhibit strong robustness when facing part of classes (easy class), but weak robustness when facing others (hard class). In this paper, we give an in-depth analysis of the potential factors and argue that the smoothness degree of samples' soft labels for different classes (i.e., hard class or easy class) will affect the robust fairness of DNN models from both empirical observation and theoretical analysis. Based on the above finding, we propose an Anti-Bias Soft Label Distillation (ABSLD) method to mitigate the adversarial robust fairness problem within the framework of Knowledge Distillation (KD). Specifically, ABSLD adaptively reduces the student's error risk gap between different classes to achieve fairness by adjusting the class-wise smoothness degree of samples' soft labels during the training process, and the smoothness degree of soft labels is controlled by assigning different temperatures in KD to different classes. Extensive experiments demonstrate that ABSLD outperforms state-of-the-art AT, ARD, and robust fairness methods in terms of overall performance of robustness and fairness.
Authors: Sheng Jin, Shuhuai Li, Tong Li, Wentao Liu, Chen Qian, Ping Luo
Human-centric perception (e.g. pedetrian detection, segmentation, pose estimation, and attribute analysis) is a long-standing problem for computer vision. This paper introduces a unified and versatile framework (HQNet) for single-stage multi-person multi-task human-centric perception (HCP). Our approach centers on learning a unified human query representation, denoted as Human Query, which captures intricate instance-level features for individual persons and disentangles complex multi-person scenarios. Although different HCP tasks have been well-studied individually, single-stage multi-task learning of HCP tasks has not been fully exploited in the literature due to the absence of a comprehensive benchmark dataset. To address this gap, we propose COCO-UniHuman benchmark dataset to enable model development and comprehensive evaluation. Experimental results demonstrate the proposed method's state-of-the-art performance among multi-task HCP models and its competitive performance compared to task-specific HCP models. Moreover, our experiments underscore Human Query's adaptability to new HCP tasks, thus demonstrating its robust generalization capability. Codes and data will be publicly accessible.
Authors: Kwang-Hyun Uhm, Hyunjun Cho, Zhixin Xu, Seohoon Lim, Seung-Won Jung, Sung-Hoo Hong, Sung-Jea Ko
In 2023, it is estimated that 81,800 kidney cancer cases will be newly diagnosed, and 14,890 people will die from this cancer in the United States. Preoperative dynamic contrast-enhanced abdominal computed tomography (CT) is often used for detecting lesions. However, there exists inter-observer variability due to subtle differences in the imaging features of kidney and kidney tumors. In this paper, we explore various 3D U-Net training configurations and effective post-processing strategies for accurate segmentation of kidneys, cysts, and kidney tumors in CT images. We validated our model on the dataset of the 2023 Kidney and Kidney Tumor Segmentation (KiTS23) challenge. Our method took second place in the final ranking of the KiTS23 challenge on unseen test data with an average Dice score of 0.820 and an average Surface Dice of 0.712.
Authors: Hao Zhang, Fang Li, Lu Qi, Ming-Hsuan Yang, Narendra Ahuja
Addressing Out-Of-Distribution (OOD) Segmentation and Zero-Shot Semantic Segmentation (ZS3) is challenging, necessitating segmenting unseen classes. Existing strategies adapt the class-agnostic Mask2Former (CA-M2F) tailored to specific tasks. However, these methods cater to singular tasks, demand training from scratch, and we demonstrate certain deficiencies in CA-M2F, which affect performance. We propose the Class-Agnostic Structure-Constrained Learning (CSL), a plug-in framework that can integrate with existing methods, thereby embedding structural constraints and achieving performance gain, including the unseen, specifically OOD, ZS3, and domain adaptation (DA) tasks. There are two schemes for CSL to integrate with existing methods (1) by distilling knowledge from a base teacher network, enforcing constraints across training and inference phrases, or (2) by leveraging established models to obtain per-pixel distributions without retraining, appending constraints during the inference phase. We propose soft assignment and mask split methodologies that enhance OOD object segmentation. Empirical evaluations demonstrate CSL's prowess in boosting the performance of existing algorithms spanning OOD segmentation, ZS3, and DA segmentation, consistently transcending the state-of-art across all three tasks.
Authors: Junzhe Lu, Jing Lin, Hongkun Dou, Yulun Zhang, Yue Deng, Haoqian Wang
Modeling human pose is a cornerstone in applications from human-robot interaction to augmented reality, yet crafting a robust human pose prior remains a challenge due to biomechanical constraints and diverse human movements. Traditional priors like VAEs and NDFs often fall short in realism and generalization, especially in extreme conditions such as unseen noisy poses. To address these issues, we introduce DPoser, a robust and versatile human pose prior built upon diffusion models. Designed with optimization frameworks, DPoser seamlessly integrates into various pose-centric applications, including human mesh recovery, pose completion, and motion denoising. Specifically, by formulating these tasks as inverse problems, we employ variational diffusion sampling for efficient solving. Furthermore, acknowledging the disparity between the articulated poses we focus on and structured images in previous research, we propose a truncated timestep scheduling to boost performance on downstream tasks. Our exhaustive experiments demonstrate DPoser's superiority over existing state-of-the-art pose priors across multiple tasks.
Authors: Kwang-Hyun Uhm, Seung-Won Jung, Moon Hyung Choi, Sung-Hoo Hong, Sung-Jea Ko
Multi-phase CT is widely adopted for the diagnosis of kidney cancer due to the complementary information among phases. However, the complete set of multi-phase CT is often not available in practical clinical applications. In recent years, there have been some studies to generate the missing modality image from the available data. Nevertheless, the generated images are not guaranteed to be effective for the diagnosis task. In this paper, we propose a unified framework for kidney cancer diagnosis with incomplete multi-phase CT, which simultaneously recovers missing CT images and classifies cancer subtypes using the completed set of images. The advantage of our framework is that it encourages a synthesis model to explicitly learn to generate missing CT phases that are helpful for classifying cancer subtypes. We further incorporate lesion segmentation network into our framework to exploit lesion-level features for effective cancer classification in the whole CT volumes. The proposed framework is based on fully 3D convolutional neural networks to jointly optimize both synthesis and classification of 3D CT volumes. Extensive experiments on both in-house and external datasets demonstrate the effectiveness of our framework for the diagnosis with incomplete data compared with state-of-the-art baselines. In particular, cancer subtype classification using the completed CT data by our method achieves higher performance than the classification using the given incomplete data.
Authors: Zhiling Ye, LiangGuo Zhang, Dingheng Zeng, Quan Lu, Ning Jiang
Dynamic NeRFs have recently garnered growing attention for 3D talking portrait synthesis. Despite advances in rendering speed and visual quality, challenges persist in enhancing efficiency and effectiveness. We present R2-Talker, an efficient and effective framework enabling realistic real-time talking head synthesis. Specifically, using multi-resolution hash grids, we introduce a novel approach for encoding facial landmarks as conditional features. This approach losslessly encodes landmark structures as conditional features, decoupling input diversity, and conditional spaces by mapping arbitrary landmarks to a unified feature space. We further propose a scheme of progressive multilayer conditioning in the NeRF rendering pipeline for effective conditional feature fusion. Our new approach has the following advantages as demonstrated by extensive experiments compared with the state-of-the-art works: 1) The lossless input encoding enables acquiring more precise features, yielding superior visual quality. The decoupling of inputs and conditional spaces improves generalizability. 2) The fusing of conditional features and MLP outputs at each MLP layer enhances conditional impact, resulting in more accurate lip synthesis and better visual quality. 3) It compactly structures the fusion of conditional features, significantly enhancing computational efficiency.
Authors: Chaoquan Jiang, Jinqiang Wang, Rui Hu, Jitao Sang
Vision models with high overall accuracy often exhibit systematic errors in specific scenarios, posing potential serious safety concerns. Diagnosing bugs of vision models is gaining increased attention, however traditional diagnostic approaches require annotation efforts (\eg rich metadata accompanying each samples of CelebA). To address this issue,We propose a language-assisted diagnostic method that uses texts instead of images to diagnose bugs in vision models based on multi-modal models (\eg CLIP). Our approach connects the embedding space of CLIP with the buggy vision model to be diagnosed; meanwhile, utilizing a shared classifier and the cross-modal transferability of embedding space from CLIP, the text-branch of CLIP become a proxy model to find bugs in the buggy model. The proxy model can classify texts paired with images. During the diagnosis, a Large Language Model (LLM) is employed to obtain task-relevant corpora, and this corpora is used to extract keywords. Descriptions constructed with templates containing these keywords serve as input text to probe errors in the proxy model. Finally, we validate the ability to diagnose existing visual models using language on the Waterbirds and CelebA datasets, we can identify bugs comprehensible to human experts, uncovering not only known bugs but also previously unknown ones.
Authors: Jifeng Dai, Yi Li, Kaiming He, Jian Sun
We present region-based, fully convolutional networks for accurate and efficient object detection. In contrast to previous region-based detectors such as Fast/Faster R-CNN that apply a costly per-region subnetwork hundreds of times, our region-based detector is fully convolutional with almost all computation shared on the entire image. To achieve this goal, we propose position-sensitive score maps to address a dilemma between translation-invariance in image classification and translation-variance in object detection. Our method can thus naturally adopt fully convolutional image classifier backbones, such as the latest Residual Networks (ResNets), for object detection. We show competitive results on the PASCAL VOC datasets (e.g., 83.6% mAP on the 2007 set) with the 101-layer ResNet. Meanwhile, our result is achieved at a test-time speed of 170ms per image, 2.5-20x faster than the Faster R-CNN counterpart. Code is made publicly available at: https://github.com/daijifeng001/r-fcn
Authors: Joshua Ball
Convolutional neural networks (CNNs) have been successful in representing the fully-connected inferencing ability perceived to be seen in the human brain: they take full advantage of the hierarchy-style patterns commonly seen in complex data and develop more patterns using simple features. Countless implementations of CNNs have shown how strong their ability is to learn these complex patterns, particularly in the realm of image classification. However, the cost of getting a high performance CNN to a so-called "state of the art" level is computationally costly. Even when using transfer learning, which utilize the very deep layers from models such as MobileNetV2, CNNs still take a great amount of time and resources. Linear discriminant analysis (LDA), a generalization of Fisher's linear discriminant, can be implemented in a multi-class classification method to increase separability of class features while not needing a high performance system to do so for image classification. Similarly, we also believe LDA has great promise in performing well. In this paper, we discuss our process of developing a robust CNN for food classification as well as our effective implementation of multi-class LDA and prove that (1) CNN is superior to LDA for image classification and (2) why LDA should not be left out of the races for image classification, particularly for binary cases.
Authors: Weicai Ye, Xinyue Lan, Ge Su, Hujun Bao, Zhaopeng Cui, Guofeng Zhang
Video Panoptic Segmentation (VPS) aims to generate coherent panoptic segmentation and track the identities of all pixels across video frames. Existing methods predominantly utilize the trained instance embedding to keep the consistency of panoptic segmentation. However, they inevitably struggle to cope with the challenges of small objects, similar appearance but inconsistent identities, occlusion, and strong instance contour deformations. To address these problems, we present HybridTracker, a lightweight and joint tracking model attempting to eliminate the limitations of the single tracker. HybridTracker performs pixel tracker and instance tracker in parallel to obtain the association matrices, which are fused into a matching matrix. In the instance tracker, we design a differentiable matching layer, ensuring the stability of inter-frame matching. In the pixel tracker, we compute the dice coefficient of the same instance of different frames given the estimated optical flow, forming the Intersection Over Union (IoU) matrix. We additionally propose mutual check and temporal consistency constraints during inference to settle the occlusion and contour deformation challenges. Comprehensive experiments show that HybridTracker achieves superior performance than state-of-the-art methods on Cityscapes-VPS and VIPER datasets.
Authors: Kang Liao, Xiangyu Xu, Chunyu Lin, Wenqi Ren, Yunchao Wei, Yao Zhao
Image outpainting gains increasing attention since it can generate the complete scene from a partial view, providing a valuable solution to construct {360\textdegree} panoramic images. As image outpainting suffers from the intrinsic issue of unidirectional completion flow, previous methods convert the original problem into inpainting, which allows a bidirectional flow. However, we find that inpainting has its own limitations and is inferior to outpainting in certain situations. The question of how they may be combined for the best of both has as yet remained under-explored. In this paper, we provide a deep analysis of the differences between inpainting and outpainting, which essentially depends on how the source pixels contribute to the unknown regions under different spatial arrangements. Motivated by this analysis, we present a Cylin-Painting framework that involves meaningful collaborations between inpainting and outpainting and efficiently fuses the different arrangements, with a view to leveraging their complementary benefits on a seamless cylinder. Nevertheless, straightforwardly applying the cylinder-style convolution often generates visually unpleasing results as it discards important positional information. To address this issue, we further present a learnable positional embedding strategy to incorporate the missing component of positional encoding into the cylinder convolution, which significantly improves the panoramic results. It is noted that while developed for image outpainting, the proposed algorithm can be effectively extended to other panoramic vision tasks, such as object detection, depth estimation, and image super-resolution. Code will be made available at \url{https://github.com/KangLiao929/Cylin-Painting}.
Authors: Rui-Jie Zhu, Qihang Zhao, Tianjing Zhang, Haoyu Deng, Yule Duan, Malu Zhang, Liang-Jian Deng
Spiking Neural Networks (SNNs) are attracting widespread interest due to their biological plausibility, energy efficiency, and powerful spatio-temporal information representation ability. Given the critical role of attention mechanisms in enhancing neural network performance, the integration of SNNs and attention mechanisms exhibits potential to deliver energy-efficient and high-performance computing paradigms. We present a novel Temporal-Channel Joint Attention mechanism for SNNs, referred to as TCJA-SNN. The proposed TCJA-SNN framework can effectively assess the significance of spike sequence from both spatial and temporal dimensions. More specifically, our essential technical contribution lies on: 1) We employ the squeeze operation to compress the spike stream into an average matrix. Then, we leverage two local attention mechanisms based on efficient 1D convolutions to facilitate comprehensive feature extraction at the temporal and channel levels independently. 2) We introduce the Cross Convolutional Fusion (CCF) layer as a novel approach to model the inter-dependencies between the temporal and channel scopes. This layer breaks the independence of these two dimensions and enables the interaction between features. Experimental results demonstrate that the proposed TCJA-SNN outperforms SOTA by up to 15.7% accuracy on standard static and neuromorphic datasets, including Fashion-MNIST, CIFAR10-DVS, N-Caltech 101, and DVS128 Gesture. Furthermore, we apply the TCJA-SNN framework to image generation tasks by leveraging a variation autoencoder. To the best of our knowledge, this study is the first instance where the SNN-attention mechanism has been employed for image classification and generation tasks. Notably, our approach has achieved SOTA performance in both domains, establishing a significant advancement in the field. Codes are available at https://github.com/ridgerchu/TCJA.
Authors: Kenan Morani, Esra Kaya Ayana, Devrim Unay
The significance of efficient and accurate diagnosis amidst the unique challenges posed by the COVID-19 pandemic underscores the urgency for innovative approaches. In response to these challenges, we propose a transfer learning-based approach using a recently annotated Computed Tomography (CT) image database. While many approaches propose an intensive data preproseccing and/or complex model architecture, our method focusses on offering an efficient solution with minimal manual engineering. Specifically, we investigate the suitability of a modified Xception model for COVID-19 detection. The method involves adapting a pre-trained Xception model, incorporating both the architecture and pre-trained weights from ImageNet. The output of the model was designed to take the final diagnosis decisions. The training utilized 128 batch sizes and 224x224 input image dimensions, downsized from standard 512x512. No further da processing was performed on the input data. Evaluation is conducted on the 'COV19-CT-DB' CT image dataset, containing labeled COVID-19 and non-COVID-19 cases. Results reveal the method's superiority in accuracy, precision, recall, and macro F1 score on the validation subset, outperforming VGG-16 transfer model and thus offering enhanced precision with fewer parameters. Furthermore, when compared to alternative methods for the COV19-CT-DB dataset, our approach exceeds the baseline approach and other alternatives on the same dataset. Finally, the adaptability of the modified Xception trasnfer learning-based model to the unique features of the COV19-CT-DB dataset showcases its potential as a robust tool for enhanced COVID-19 diagnosis from CT images.
Authors: Xingguang Zhang, Zhiyuan Mao, Nicholas Chimitt, Stanley H. Chan
Restoring images distorted by atmospheric turbulence is a ubiquitous problem in long-range imaging applications. While existing deep-learning-based methods have demonstrated promising results in specific testing conditions, they suffer from three limitations: (1) lack of generalization capability from synthetic training data to real turbulence data; (2) failure to scale, hence causing memory and speed challenges when extending the idea to a large number of frames; (3) lack of a fast and accurate simulator to generate data for training neural networks. In this paper, we introduce the turbulence mitigation transformer (TMT) that explicitly addresses these issues. TMT brings three contributions: Firstly, TMT explicitly uses turbulence physics by decoupling the turbulence degradation and introducing a multi-scale loss for removing distortion, thus improving effectiveness. Secondly, TMT presents a new attention module along the temporal axis to extract extra features efficiently, thus improving memory and speed. Thirdly, TMT introduces a new simulator based on the Fourier sampler, temporal correlation, and flexible kernel size, thus improving our capability to synthesize better training data. TMT outperforms state-of-the-art video restoration models, especially in generalizing from synthetic to real turbulence data. Code, videos, and datasets are available at \href{https://xg416.github.io/TMT}{https://xg416.github.io/TMT}.
Authors: Yuwei Sun, Hideya Ochiai
Visual Question Answering (VQA) based on multi-modal data facilitates real-life applications such as home robots and medical diagnoses. One significant challenge is to devise a robust decentralized learning framework for various client models where centralized data collection is refrained due to confidentiality concerns. This work aims to tackle privacy-preserving VQA by decoupling a multi-modal model into representation modules and a contrastive module and leveraging inter-module gradients sharing and inter-client weight sharing. To this end, we propose Bidirectional Contrastive Split Learning (BiCSL) to train a global multi-modal model on the entire data distribution of decentralized clients. We employ the contrastive loss that enables a more efficient self-supervised learning of decentralized modules. Comprehensive experiments are conducted on the VQA-v2 dataset based on five SOTA VQA models, demonstrating the effectiveness of the proposed method. Furthermore, we inspect BiCSL's robustness against a dual-key backdoor attack on VQA. Consequently, BiCSL shows much better robustness to the multi-modal adversarial attack compared to the centralized learning method, which provides a promising approach to decentralized multi-modal learning.
Authors: Se-In Jang, Tinsu Pan, Ye Li, Pedram Heidari, Junyu Chen, Quanzheng Li, Kuang Gong
Position emission tomography (PET) is widely used in clinics and research due to its quantitative merits and high sensitivity, but suffers from low signal-to-noise ratio (SNR). Recently convolutional neural networks (CNNs) have been widely used to improve PET image quality. Though successful and efficient in local feature extraction, CNN cannot capture long-range dependencies well due to its limited receptive field. Global multi-head self-attention (MSA) is a popular approach to capture long-range information. However, the calculation of global MSA for 3D images has high computational costs. In this work, we proposed an efficient spatial and channel-wise encoder-decoder transformer, Spach Transformer, that can leverage spatial and channel information based on local and global MSAs. Experiments based on datasets of different PET tracers, i.e., $^{18}$F-FDG, $^{18}$F-ACBC, $^{18}$F-DCFPyL, and $^{68}$Ga-DOTATATE, were conducted to evaluate the proposed framework. Quantitative results show that the proposed Spach Transformer framework outperforms state-of-the-art deep learning architectures. Our codes are available at https://github.com/sijang/SpachTransformer
Authors: Yiqi Jin, Ziyu Zhu, Tongda Xu, Yuhuan Lin, Yan Wang
Recently, deep learning methods have shown promising results in point cloud compression. For octree-based point cloud compression, previous works show that the information of ancestor nodes and sibling nodes are equally important for predicting current node. However, those works either adopt insufficient context or bring intolerable decoding complexity (e.g. >600s). To address this problem, we propose a sufficient yet efficient context model and design an efficient deep learning codec for point clouds. Specifically, we first propose a window-constrained multi-group coding strategy to exploit the autoregressive context while maintaining decoding efficiency. Then, we propose a dual transformer architecture to utilize the dependency of current node on its ancestors and siblings. We also propose a random-masking pre-train method to enhance our model. Experimental results show that our approach achieves state-of-the-art performance for both lossy and lossless point cloud compression. Moreover, our multi-group coding strategy saves 98% decoding time compared with previous octree-based compression method.
Authors: Krystian Chachuła, Jakub Łyskawa, Bartłomiej Olber, Piotr Frątczak, Adam Popowicz, Krystian Radlak
The quality of training datasets for deep neural networks is a key factor contributing to the accuracy of resulting models. This effect is amplified in difficult tasks such as object detection. Dealing with errors in datasets is often limited to accepting that some fraction of examples are incorrect, estimating their confidence, and either assigning appropriate weights or ignoring uncertain ones during training. In this work, we propose a different approach. We introduce the Confident Learning for Object Detection (CLOD) algorithm for assessing the quality of each label in object detection datasets, identifying missing, spurious, mislabeled, and mislocated bounding boxes and suggesting corrections. By focusing on finding incorrect examples in the training datasets, we can eliminate them at the root. Suspicious bounding boxes can be reviewed to improve the quality of the dataset, leading to better models without further complicating their already complex architectures. The proposed method is able to point out nearly 80% of artificially disturbed bounding boxes with a false positive rate below 0.1. Cleaning the datasets by applying the most confident automatic suggestions improved mAP scores by 16% to 46%, depending on the dataset, without any modifications to the network architectures. This approach shows promising potential in rectifying state-of-the-art object detection datasets.
Authors: Xiangxiang Chu, Liang Li, Bo Zhang
The tradeoff between performance and inference speed is critical for practical applications. Architecture reparameterization obtains better tradeoffs and it is becoming an increasingly popular ingredient in modern convolutional neural networks. Nonetheless, its quantization performance is usually too poor to deploy (more than 20% top-1 accuracy drop on ImageNet) when INT8 inference is desired. In this paper, we dive into the underlying mechanism of this failure, where the original design inevitably enlarges quantization error. We propose a simple, robust, and effective remedy to have a quantization-friendly structure that also enjoys reparameterization benefits. Our method greatly bridges the gap between INT8 and FP32 accuracy for RepVGG. Without bells and whistles, the top-1 accuracy drop on ImageNet is reduced within 2% by standard post-training quantization. Moreover, our method also achieves similar FP32 performance as RepVGG. Extensive experiments on detection and semantic segmentation tasks verify its generalization.
Authors: Xiaoshui Huang, Zhou Huang, Sheng Li, Wentao Qu, Tong He, Yuenan Hou, Yifan Zuo, Wanli Ouyang
The pretrain-finetune paradigm has achieved great success in NLP and 2D image fields because of the high-quality representation ability and transferability of their pretrained models. However, pretraining such a strong model is difficult in the 3D point cloud field due to the limited amount of point cloud sequences. This paper introduces \textbf{E}fficient \textbf{P}oint \textbf{C}loud \textbf{L}earning (EPCL), an effective and efficient point cloud learner for directly training high-quality point cloud models with a frozen CLIP transformer. Our EPCL connects the 2D and 3D modalities by semantically aligning the image features and point cloud features without paired 2D-3D data. Specifically, the input point cloud is divided into a series of local patches, which are converted to token embeddings by the designed point cloud tokenizer. These token embeddings are concatenated with a task token and fed into the frozen CLIP transformer to learn point cloud representation. The intuition is that the proposed point cloud tokenizer projects the input point cloud into a unified token space that is similar to the 2D images. Comprehensive experiments on 3D detection, semantic segmentation, classification and few-shot learning demonstrate that the CLIP transformer can serve as an efficient point cloud encoder and our method achieves promising performance on both indoor and outdoor benchmarks. In particular, performance gains brought by our EPCL are $\textbf{19.7}$ AP$_{50}$ on ScanNet V2 detection, $\textbf{4.4}$ mIoU on S3DIS segmentation and $\textbf{1.2}$ mIoU on SemanticKITTI segmentation compared to contemporary pretrained models. Code is available at \url{https://github.com/XiaoshuiHuang/EPCL}.
Authors: Zongsheng Yue, Chen Change Loy
While deep learning-based methods for blind face restoration have achieved unprecedented success, they still suffer from two major limitations. First, most of them deteriorate when facing complex degradations out of their training data. Second, these methods require multiple constraints, e.g., fidelity, perceptual, and adversarial losses, which require laborious hyper-parameter tuning to stabilize and balance their influences. In this work, we propose a novel method named DifFace that is capable of coping with unseen and complex degradations more gracefully without complicated loss designs. The key of our method is to establish a posterior distribution from the observed low-quality (LQ) image to its high-quality (HQ) counterpart. In particular, we design a transition distribution from the LQ image to the intermediate state of a pre-trained diffusion model and then gradually transmit from this intermediate state to the HQ target by recursively applying a pre-trained diffusion model. The transition distribution only relies on a restoration backbone that is trained with $L_2$ loss on some synthetic data, which favorably avoids the cumbersome training process in existing methods. Moreover, the transition distribution can contract the error of the restoration backbone and thus makes our method more robust to unknown degradations. Comprehensive experiments show that DifFace is superior to current state-of-the-art methods, especially in cases with severe degradations. Code and model are available at https://github.com/zsyOAOA/DifFace.
Authors: Asma Ahmed Hashmi, Aigerim Zhumabayeva, Nikita Kotelevskii, Artem Agafonov, Mohammad Yaqub, Maxim Panov, Martin Takáč
The success of Deep Neural Network (DNN) models significantly depends on the quality of provided annotations. In medical image segmentation, for example, having multiple expert annotations for each data point is common to minimize subjective annotation bias. Then, the goal of estimation is to filter out the label noise and recover the ground-truth masks, which are not explicitly given. This paper proposes a probabilistic model for noisy observations that allows us to build a confident classification and segmentation models. To accomplish it, we explicitly model label noise and introduce a new information-based regularization that pushes the network to recover the ground-truth labels. In addition, for segmentation task we adjust the loss function by prioritizing learning in high-confidence regions where all the annotators agree on labeling. We evaluate the proposed method on a series of classification tasks such as noisy versions of MNIST, CIFAR-10, Fashion-MNIST datasets as well as CIFAR-10N, which is real-world dataset with noisy human annotations. Additionally, for segmentation task, we consider several medical imaging datasets, such as, LIDC and RIGA that reflect real-world inter-variability among multiple annotators. Our experiments show that our algorithm outperforms state-of-the-art solutions for the considered classification and segmentation problems.
Authors: Jerone T. A. Andrews, Dora Zhao, William Thong, Apostolos Modas, Orestis Papakyriakopoulos, Alice Xiang
Human-centric computer vision (HCCV) data curation practices often neglect privacy and bias concerns, leading to dataset retractions and unfair models. HCCV datasets constructed through nonconsensual web scraping lack crucial metadata for comprehensive fairness and robustness evaluations. Current remedies are post hoc, lack persuasive justification for adoption, or fail to provide proper contextualization for appropriate application. Our research focuses on proactive, domain-specific recommendations, covering purpose, privacy and consent, and diversity, for curating HCCV evaluation datasets, addressing privacy and bias concerns. We adopt an ante hoc reflective perspective, drawing from current practices, guidelines, dataset withdrawals, and audits, to inform our considerations and recommendations.
Authors: Paul Pu Liang, Yun Cheng, Xiang Fan, Chun Kai Ling, Suzanne Nie, Richard Chen, Zihao Deng, Nicholas Allen, Randy Auerbach, Faisal Mahmood, Ruslan Salakhutdinov, Louis-Philippe Morency
The recent explosion of interest in multimodal applications has resulted in a wide selection of datasets and methods for representing and integrating information from different modalities. Despite these empirical advances, there remain fundamental research questions: How can we quantify the interactions that are necessary to solve a multimodal task? Subsequently, what are the most suitable multimodal models to capture these interactions? To answer these questions, we propose an information-theoretic approach to quantify the degree of redundancy, uniqueness, and synergy relating input modalities with an output task. We term these three measures as the PID statistics of a multimodal distribution (or PID for short), and introduce two new estimators for these PID statistics that scale to high-dimensional distributions. To validate PID estimation, we conduct extensive experiments on both synthetic datasets where the PID is known and on large-scale multimodal benchmarks where PID estimations are compared with human annotations. Finally, we demonstrate their usefulness in (1) quantifying interactions within multimodal datasets, (2) quantifying interactions captured by multimodal models, (3) principled approaches for model selection, and (4) three real-world case studies engaging with domain experts in pathology, mood prediction, and robotic perception where our framework helps to recommend strong multimodal models for each application.
Authors: David Svitov, Dmitrii Gudkov, Renat Bashirov, Victor Lempitsky
We present DINAR, an approach for creating realistic rigged fullbody avatars from single RGB images. Similarly to previous works, our method uses neural textures combined with the SMPL-X body model to achieve photo-realistic quality of avatars while keeping them easy to animate and fast to infer. To restore the texture, we use a latent diffusion model and show how such model can be trained in the neural texture space. The use of the diffusion model allows us to realistically reconstruct large unseen regions such as the back of a person given the frontal view. The models in our pipeline are trained using 2D images and videos only. In the experiments, our approach achieves state-of-the-art rendering quality and good generalization to new poses and viewpoints. In particular, the approach improves state-of-the-art on the SnapshotPeople public benchmark.
Authors: Duolikun Danier, Fan Zhang, David Bull
Existing works on video frame interpolation (VFI) mostly employ deep neural networks that are trained by minimizing the L1, L2, or deep feature space distance (e.g. VGG loss) between their outputs and ground-truth frames. However, recent works have shown that these metrics are poor indicators of perceptual VFI quality. Towards developing perceptually-oriented VFI methods, in this work we propose latent diffusion model-based VFI, LDMVFI. This approaches the VFI problem from a generative perspective by formulating it as a conditional generation problem. As the first effort to address VFI using latent diffusion models, we rigorously benchmark our method on common test sets used in the existing VFI literature. Our quantitative experiments and user study indicate that LDMVFI is able to interpolate video content with favorable perceptual quality compared to the state of the art, even in the high-resolution regime. Our code is available at https://github.com/danier97/LDMVFI.
Authors: Gongpei Zhao, Tao Wang, Yidong Li, Yi Jin, Congyan Lang, Haibin Ling
Backpropagation algorithm has been widely used as a mainstream learning procedure for neural networks in the past decade, and has played a significant role in the development of deep learning. However, there exist some limitations associated with this algorithm, such as getting stuck in local minima and experiencing vanishing/exploding gradients, which have led to questions about its biological plausibility. To address these limitations, alternative algorithms to backpropagation have been preliminarily explored, with the Forward-Forward (FF) algorithm being one of the most well-known. In this paper we propose a new learning framework for neural networks, namely Cascaded Forward (CaFo) algorithm, which does not rely on BP optimization as that in FF. Unlike FF, our framework directly outputs label distributions at each cascaded block, which does not require generation of additional negative samples and thus leads to a more efficient process at both training and testing. Moreover, in our framework each block can be trained independently, so it can be easily deployed into parallel acceleration systems. The proposed method is evaluated on four public image classification benchmarks, and the experimental results illustrate significant improvement in prediction accuracy in comparison with the baseline.
Authors: Renat Bashirov, Alexey Larionov, Evgeniya Ustinova, Mikhail Sidorenko, David Svitov, Ilya Zakharkin, Victor Lempitsky
We present a system to create Mobile Realistic Fullbody (MoRF) avatars. MoRF avatars are rendered in real-time on mobile devices, learned from monocular videos, and have high realism. We use SMPL-X as a proxy geometry and render it with DNR (neural texture and image-2-image network). We improve on prior work, by overfitting per-frame warping fields in the neural texture space, allowing to better align the training signal between different frames. We also refine SMPL-X mesh fitting procedure to improve the overall avatar quality. In the comparisons to other monocular video-based avatar systems, MoRF avatars achieve higher image sharpness and temporal consistency. Participants of our user study also preferred avatars generated by MoRF.
Authors: Charalambos Tzamos, Daniel Barath, Torsten Sattler, Zuzana Kukelova
We study challenging problems of estimating the relative pose of three cameras and propose novel efficient solutions to (1) the notoriously difficult configuration of four points in three calibrated views, known as the 4p3v problem, and (2) to the previously unsolved configuration of four points in three cameras with unknown shared focal length, i.e., the 4p3vf problem. Our solutions are based on the simple idea of generating one or two additional virtual point correspondences in two views by using the information from the locations of the four input correspondences in the three views. We generate such correspondences using either a very simple and efficient strategy where the new points are the mean points of three corresponding input points or using a simple neural network. The new solvers are efficient and easy to implement since they are based on existing efficient minimal solvers, i.e., the well-known 5-point and 6-point relative pose solvers and the P3P solver. Our solvers achieve state-of-the-art results on real data.
Authors: Yanda Li, Zilong Huang, Gang Yu, Ling Chen, Yunchao Wei, Jianbo Jiao
Image matting requires high-quality pixel-level human annotations to support the training of a deep model in recent literature. Whereas such annotation is costly and hard to scale, significantly holding back the development of the research. In this work, we make the first attempt towards addressing this problem, by proposing a self-supervised pre-training approach that can leverage infinite numbers of data to boost the matting performance. The pre-training task is designed in a similar manner as image matting, where random trimap and alpha matte are generated to achieve an image disentanglement objective. The pre-trained model is then used as an initialisation of the downstream matting task for fine-tuning. Extensive experimental evaluations show that the proposed approach outperforms both the state-of-the-art matting methods and other alternative self-supervised initialisation approaches by a large margin. We also show the robustness of the proposed approach over different backbone architectures. Our project page is available at https://crystraldo.github.io/dpt_mat/.
Authors: Cheng Sun, Guangyan Cai, Zhengqin Li, Kai Yan, Cheng Zhang, Carl Marshall, Jia-Bin Huang, Shuang Zhao, Zhao Dong
Reconstructing the shape and spatially varying surface appearances of a physical-world object as well as its surrounding illumination based on 2D images (e.g., photographs) of the object has been a long-standing problem in computer vision and graphics. In this paper, we introduce an accurate and highly efficient object reconstruction pipeline combining neural based object reconstruction and physics-based inverse rendering (PBIR). Our pipeline firstly leverages a neural SDF based shape reconstruction to produce high-quality but potentially imperfect object shape. Then, we introduce a neural material and lighting distillation stage to achieve high-quality predictions for material and illumination. In the last stage, initialized by the neural predictions, we perform PBIR to refine the initial results and obtain the final high-quality reconstruction of object shape, material, and illumination. Experimental results demonstrate our pipeline significantly outperforms existing methods quality-wise and performance-wise.
Authors: Xu Yang, Jiawei Peng, Zihua Wang, Haiyang Xu, Qinghao Ye, Chenliang Li, Songfang Huang, Fei Huang, Zhangzikang Li, Yu Zhang
We propose to Transform Scene Graphs (TSG) into more descriptive captions. In TSG, we apply multi-head attention (MHA) to design the Graph Neural Network (GNN) for embedding scene graphs. After embedding, different graph embeddings contain diverse specific knowledge for generating the words with different part-of-speech, e.g., object/attribute embedding is good for generating nouns/adjectives. Motivated by this, we design a Mixture-of-Expert (MOE)-based decoder, where each expert is built on MHA, for discriminating the graph embeddings to generate different kinds of words. Since both the encoder and decoder are built based on the MHA, as a result, we construct a homogeneous encoder-decoder unlike the previous heterogeneous ones which usually apply Fully-Connected-based GNN and LSTM-based decoder. The homogeneous architecture enables us to unify the training configuration of the whole model instead of specifying different training strategies for diverse sub-networks as in the heterogeneous pipeline, which releases the training difficulty. Extensive experiments on the MS-COCO captioning benchmark validate the effectiveness of our TSG. The code is in: https://github.com/GaryJiajia/TSG.
Authors: Fa-Ting Hong, Li Shen, Dan Xu
Predominant techniques on talking head generation largely depend on 2D information, including facial appearances and motions from input face images. Nevertheless, dense 3D facial geometry, such as pixel-wise depth, plays a critical role in constructing accurate 3D facial structures and suppressing complex background noises for generation. However, dense 3D annotations for facial videos is prohibitively costly to obtain. In this work, firstly, we present a novel self-supervised method for learning dense 3D facial geometry (ie, depth) from face videos, without requiring camera parameters and 3D geometry annotations in training. We further propose a strategy to learn pixel-level uncertainties to perceive more reliable rigid-motion pixels for geometry learning. Secondly, we design an effective geometry-guided facial keypoint estimation module, providing accurate keypoints for generating motion fields. Lastly, we develop a 3D-aware cross-modal (ie, appearance and depth) attention mechanism, which can be applied to each generation layer, to capture facial geometries in a coarse-to-fine manner. Extensive experiments are conducted on three challenging benchmarks (ie, VoxCeleb1, VoxCeleb2, and HDTF). The results demonstrate that our proposed framework can generate highly realistic-looking reenacted talking videos, with new state-of-the-art performances established on these benchmarks. The codes and trained models are publicly available on the GitHub project page at https://github.com/harlanhong/CVPR2022-DaGAN
Authors: Hassan Akbari, Dan Kondratyuk, Yin Cui, Rachel Hornung, Huisheng Wang, Hartwig Adam
We present Integrated Multimodal Perception (IMP), a simple and scalable multimodal multi-task training and modeling approach. IMP integrates multimodal inputs including image, video, text, and audio into a single Transformer encoder with minimal modality-specific components. IMP makes use of a novel design that combines Alternating Gradient Descent (AGD) and Mixture-of-Experts (MoE) for efficient model and task scaling. We conduct extensive empirical studies and reveal the following key insights: 1) Performing gradient descent updates by alternating on diverse modalities, loss functions, and tasks, with varying input resolutions, efficiently improves the model. 2) Sparsification with MoE on a single modality-agnostic encoder substantially improves the performance, outperforming dense models that use modality-specific encoders or additional fusion layers and greatly mitigates the conflicts between modalities. IMP achieves competitive performance on a wide range of downstream tasks including video classification, image classification, image-text, and video-text retrieval. Most notably, we train a sparse IMP-MoE-L variant focusing on video tasks that achieves new state-of-the-art in zero-shot video classification: 77.0% on Kinetics-400, 76.8% on Kinetics-600, and 68.3% on Kinetics-700, improving the previous state-of-the-art by +5%, +6.7%, and +5.8%, respectively, while using only 15% of their total training computational cost.
Authors: Nico Montali, John Lambert, Paul Mougin, Alex Kuefler, Nick Rhinehart, Michelle Li, Cole Gulino, Tristan Emrich, Zoey Yang, Shimon Whiteson, Brandyn White, Dragomir Anguelov
Simulation with realistic, interactive agents represents a key task for autonomous vehicle software development. In this work, we introduce the Waymo Open Sim Agents Challenge (WOSAC). WOSAC is the first public challenge to tackle this task and propose corresponding metrics. The goal of the challenge is to stimulate the design of realistic simulators that can be used to evaluate and train a behavior model for autonomous driving. We outline our evaluation methodology, present results for a number of different baseline simulation agent methods, and analyze several submissions to the 2023 competition which ran from March 16, 2023 to May 23, 2023. The WOSAC evaluation server remains open for submissions and we discuss open problems for the task.
Authors: Shengqin Jiang, Yaoyu Fang, Haokui Zhang, Qingshan Liu, Yuankai Qi, Yang Yang, Peng Wang
Rehearsal-based video incremental learning often employs knowledge distillation to mitigate catastrophic forgetting of previously learned data. However, this method faces two major challenges for video task: substantial computing resources from loading teacher model and limited replay capability from performance-limited teacher model. To address these problems, we first propose a knowledge distillation-free framework for rehearsal-based video incremental learning called \textit{Teacher Agent}. Instead of loading parameter-heavy teacher networks, we introduce an agent generator that is either parameter-free or uses only a few parameters to obtain accurate and reliable soft labels. This method not only greatly reduces the computing requirement but also circumvents the problem of knowledge misleading caused by inaccurate predictions of the teacher model. Moreover, we put forward a self-correction loss which provides an effective regularization signal for the review of old knowledge, which in turn alleviates the problem of catastrophic forgetting. Further, to ensure that the samples in the memory buffer are memory-efficient and representative, we introduce a unified sampler for rehearsal-based video incremental learning to mine fixed-length key video frames. Interestingly, based on the proposed strategies, the network exhibits a high level of robustness against spatial resolution reduction when compared to the baseline. Extensive experiments demonstrate the advantages of our method, yielding significant performance improvements while utilizing only half the spatial resolution of video clips as network inputs in the incremental phases.
Authors: Junhyuk So, Jungwon Lee, Daehyun Ahn, Hyungjun Kim, Eunhyeok Park
The diffusion model has gained popularity in vision applications due to its remarkable generative performance and versatility. However, high storage and computation demands, resulting from the model size and iterative generation, hinder its use on mobile devices. Existing quantization techniques struggle to maintain performance even in 8-bit precision due to the diffusion model's unique property of temporal variation in activation. We introduce a novel quantization method that dynamically adjusts the quantization interval based on time step information, significantly improving output quality. Unlike conventional dynamic quantization techniques, our approach has no computational overhead during inference and is compatible with both post-training quantization (PTQ) and quantization-aware training (QAT). Our extensive experiments demonstrate substantial improvements in output quality with the quantized diffusion model across various datasets.
Authors: Thomas Walker, Octave Mariotti, Amir Vaxman, Hakan Bilen
We introduce Explicit Neural Surfaces (ENS), an efficient smooth surface representation that directly encodes topology with a deformation field from a known base domain. We apply this representation to reconstruct explicit surfaces from multiple views, where we use a series of neural deformation fields to progressively transform the base domain into a target shape. By using meshes as discrete surface proxies, we train the deformation fields through efficient differentiable rasterization. Using a fixed base domain allows us to have Laplace-Beltrami eigenfunctions as an intrinsic positional encoding alongside standard extrinsic Fourier features, with which our approach can capture fine surface details. Compared to implicit surfaces, ENS trains faster and has several orders of magnitude faster inference times. The explicit nature of our approach also allows higher-quality mesh extraction whilst maintaining competitive surface reconstruction performance and real-time capabilities.
Authors: Boyuan Sun, Yuqi Yang, Le Zhang, Ming-Ming Cheng, Qibin Hou
This paper presents a simple but performant semi-supervised semantic segmentation approach, called CorrMatch. Previous approaches mostly employ complicated training strategies to leverage unlabeled data but overlook the role of correlation maps in modeling the relationships between pairs of locations. We observe that the correlation maps not only enable clustering pixels of the same category easily but also contain good shape information, which previous works have omitted. Motivated by these, we aim to improve the use efficiency of unlabeled data by designing two novel label propagation strategies. First, we propose to conduct pixel propagation by modeling the pairwise similarities of pixels to spread the high-confidence pixels and dig out more. Then, we perform region propagation to enhance the pseudo labels with accurate class-agnostic masks extracted from the correlation maps. CorrMatch achieves great performance on popular segmentation benchmarks. Taking the DeepLabV3+ with ResNet-101 backbone as our segmentation model, we receive a 76%+ mIoU score on the Pascal VOC 2012 dataset with only 92 annotated images. Code is available at https://github.com/BBBBchan/CorrMatch.
Authors: Deng Yu, Chufeng Xiao, Manfred Lau, Hongbo Fu
In the process of product design and digital fabrication, the structural analysis of a designed prototype is a fundamental and essential step. However, such a step is usually invisible or inaccessible to designers at the early sketching phase. This limits the user's ability to consider a shape's physical properties and structural soundness. To bridge this gap, we introduce a novel approach Sketch2Stress that allows users to perform structural analysis of desired objects at the sketching stage. This method takes as input a 2D freehand sketch and one or multiple locations of user-assigned external forces. With the specially-designed two-branch generative-adversarial framework, it automatically predicts a normal map and a corresponding structural stress map distributed over the user-sketched underlying object. In this way, our method empowers designers to easily examine the stress sustained everywhere and identify potential problematic regions of their sketched object. Furthermore, combined with the predicted normal map, users are able to conduct a region-wise structural analysis efficiently by aggregating the stress effects of multiple forces in the same direction. Finally, we demonstrate the effectiveness and practicality of our system with extensive experiments and user studies.
Authors: Taimeng Fu, Shaoshu Su, Yiren Lu, Chen Wang
Simultaneous Localization and Mapping (SLAM) stands as one of the critical challenges in robot navigation. A SLAM system often consists of a front-end component for motion estimation and a back-end system for eliminating estimation drift. Recent advancements suggest that data-driven methods are highly effective for front-end tasks, while geometry-based methods continue to be essential in the back-end processes. However, such a decoupled paradigm between the data-driven front-end and geometry-based back-end can lead to sub-optimal performance, consequently reducing system capabilities and generalization potential. To solve this problem, we proposed a novel self-supervised imperative learning framework, named imperative SLAM (iSLAM), which fosters reciprocal correction between the front-end and back-end, thus enhancing performance without necessitating any external supervision. Specifically, we formulate the SLAM problem as a bilevel optimization so that the front-end and back-end are bidirectionally connected. As a result, the front-end model can learn global geometric knowledge obtained through pose graph optimization by back-propagating the residuals from the back-end component. We showcase the effectiveness of this new framework through an application of stereo-inertial SLAM. The experiments show that the iSLAM training strategy achieves an accuracy improvement of 22% on average over a baseline model. To the best of our knowledge, iSLAM is the first SLAM system showing that the front-end and back-end can mutually correct each other in a self-supervised manner.
Authors: Stephanie Fu, Netanel Tamir, Shobhita Sundaram, Lucy Chai, Richard Zhang, Tali Dekel, Phillip Isola
Current perceptual similarity metrics operate at the level of pixels and patches. These metrics compare images in terms of their low-level colors and textures, but fail to capture mid-level similarities and differences in image layout, object pose, and semantic content. In this paper, we develop a perceptual metric that assesses images holistically. Our first step is to collect a new dataset of human similarity judgments over image pairs that are alike in diverse ways. Critical to this dataset is that judgments are nearly automatic and shared by all observers. To achieve this we use recent text-to-image models to create synthetic pairs that are perturbed along various dimensions. We observe that popular perceptual metrics fall short of explaining our new data, and we introduce a new metric, DreamSim, tuned to better align with human perception. We analyze how our metric is affected by different visual attributes, and find that it focuses heavily on foreground objects and semantic content while also being sensitive to color and layout. Notably, despite being trained on synthetic data, our metric generalizes to real images, giving strong results on retrieval and reconstruction tasks. Furthermore, our metric outperforms both prior learned metrics and recent large vision models on these tasks.
Authors: Marcos V. Conde, Javier Vazquez-Corral, Michael S. Brown, Radu Timofte
3D lookup tables (3D LUTs) are a key component for image enhancement. Modern image signal processors (ISPs) have dedicated support for these as part of the camera rendering pipeline. Cameras typically provide multiple options for picture styles, where each style is usually obtained by applying a unique handcrafted 3D LUT. Current approaches for learning and applying 3D LUTs are notably fast, yet not so memory-efficient, as storing multiple 3D LUTs is required. For this reason and other implementation limitations, their use on mobile devices is less popular. In this work, we propose a Neural Implicit LUT (NILUT), an implicitly defined continuous 3D color transformation parameterized by a neural network. We show that NILUTs are capable of accurately emulating real 3D LUTs. Moreover, a NILUT can be extended to incorporate multiple styles into a single network with the ability to blend styles implicitly. Our novel approach is memory-efficient, controllable and can complement previous methods, including learned ISPs. Code, models and dataset available at: https://github.com/mv-lab/nilut
Authors: Raúl Ramos-Pollán, Fabio A. González
This work addresses the challenge of training supervised machine or deep learning models on orbiting platforms where we are generally constrained by limited on-board hardware capabilities and restricted uplink bandwidths to upload. We aim at enabling orbiting spacecrafts to (1) continuously train a lightweight model as it acquires imagery; and (2) receive new labels while on orbit to refine or even change the predictive task being trained. For this, we consider chip level regression tasks (i.e. predicting the vegetation percentage of a 20 km$^2$ patch) when we only have coarser label proportions, such as municipality level vegetation statistics (a municipality containing several patches). Such labels proportions have the additional advantage that usually come in tabular data and are widely available in many regions of the world and application areas. This can be framed as a Learning from Label Proportions (LLP) problem setup. LLP applied to Earth Observation (EO) data is still an emerging field and performing comparative studies in applied scenarios remains a challenge due to the lack of standardized datasets. In this work, first, we show how very simple deep learning and probabilistic methods (with {\raise.17ex\hbox{$\scriptstyle\sim$}}5K parameters) generally perform better than standard more complex ones, providing a surprising level of finer grained spatial detail when trained with much coarser label proportions. Second, we publish a set of benchmarking datasets enabling comparative LLP applied to EO, providing both fine grained labels and aggregated data according to existing administrative divisions. Finally, we show how this approach fits an on-orbit training scenario by reducing vastly both the amount of computing and the size of the labels sets. Source code is available at https://github.com/rramosp/llpeo
Authors: Youngchan Kim, Wonjoon Jin, Sunghyun Cho, Seung-Hwan Baek
Modeling the spatial radiance distribution of light rays in a scene has been extensively explored for applications, including view synthesis. Spectrum and polarization, the wave properties of light, are often neglected due to their integration into three RGB spectral bands and their non-perceptibility to human vision. However, these properties are known to encompass substantial material and geometric information about a scene. Here, we propose to model spectro-polarimetric fields, the spatial Stokes-vector distribution of any light ray at an arbitrary wavelength. We present Neural Spectro-polarimetric Fields (NeSpoF), a neural representation that models the physically-valid Stokes vector at given continuous variables of position, direction, and wavelength. NeSpoF manages inherently noisy raw measurements, showcases memory efficiency, and preserves physically vital signals - factors that are crucial for representing the high-dimensional signal of a spectro-polarimetric field. To validate NeSpoF, we introduce the first multi-view hyperspectral-polarimetric image dataset, comprised of both synthetic and real-world scenes. These were captured using our compact hyperspectral-polarimetric imaging system, which has been calibrated for robustness against system imperfections. We demonstrate the capabilities of NeSpoF on diverse scenes.
Authors: Ehsan Moradpur-Tari (1), Sergei Vlassov (1,2), Sven Oras (1,2), Mart Ernits (1), Elyad Damerchi (1), Boris Polyakovc (3), Andreas Kyritsakis (1), Veronika Zadin (1) ((1) Institute of Technology, University of Tartu, Nooruse 1, 50411 Tartu, Estonia (2) Institute of Physics, University of Tartu, W. Ostwaldi 1, 50411 Tartu, Estonia (3) Institute of Solid State Physics, University of Latvia, Kengaraga street 8, LV-1063 Riga, Latvia)
Nanoparticles in microscopy images are usually analyzed qualitatively or manually and there is a need for autonomous quantitative analysis of these objects. In this paper, we present a physics-based computational model for accurate segmentation and geometrical analysis of one-dimensional deformable overlapping objects from microscopy images. This model, named Nano1D, has four steps of preprocessing, segmentation, separating overlapped objects and geometrical measurements. The model is tested on SEM images of Ag and Au nanowire taken from different microscopes, and thermally fragmented Ag nanowires transformed into nanoparticles with different lengths, diameters, and population densities. It successfully segments and analyzes their geometrical characteristics including lengths and average diameter. The function of the algorithm is not undermined by the size, number, density, orientation and overlapping of objects in images. The main strength of the model is shown to be its ability to segment and analyze overlapping objects successfully with more than 99% accuracy, while current machine learning and computational models suffer from inaccuracy and inability to segment overlapping objects. Benefiting from a graphical user interface, Nano1D can analyze 1D nanoparticles including nanowires, nanotubes, nanorods in addition to other 1D features of microstructures like microcracks, dislocations etc.
Authors: Qi Bi, Shaodi You, Theo Gevers
Domain-generalized urban-scene semantic segmentation (USSS) aims to learn generalized semantic predictions across diverse urban-scene styles. Unlike domain gap challenges, USSS is unique in that the semantic categories are often similar in different urban scenes, while the styles can vary significantly due to changes in urban landscapes, weather conditions, lighting, and other factors. Existing approaches typically rely on convolutional neural networks (CNNs) to learn the content of urban scenes.
In this paper, we propose a Content-enhanced Mask TransFormer (CMFormer) for domain-generalized USSS. The main idea is to enhance the focus of the fundamental component, the mask attention mechanism, in Transformer segmentation models on content information. To achieve this, we introduce a novel content-enhanced mask attention mechanism. It learns mask queries from both the image feature and its down-sampled counterpart, as lower-resolution image features usually contain more robust content information and are less sensitive to style variations. These features are fused into a Transformer decoder and integrated into a multi-resolution content-enhanced mask attention learning scheme.
Extensive experiments conducted on various domain-generalized urban-scene segmentation datasets demonstrate that the proposed CMFormer significantly outperforms existing CNN-based methods for domain-generalized semantic segmentation, achieving improvements of up to 14.00\% in terms of mIoU (mean intersection over union). The source code is publicly available at \url{https://github.com/BiQiWHU/CMFormer}.
Authors: Shuo He, Lei Feng, Guowu Yang
Partial-label learning (PLL) relies on a key assumption that the true label of each training example must be in the candidate label set. This restrictive assumption may be violated in complex real-world scenarios, and thus the true label of some collected examples could be unexpectedly outside the assigned candidate label set. In this paper, we term the examples whose true label is outside the candidate label set OOC (out-of-candidate) examples, and pioneer a new PLL study to learn with OOC examples. We consider two types of OOC examples in reality, i.e., the closed-set/open-set OOC examples whose true label is inside/outside the known label space. To solve this new PLL problem, we first calculate the wooden cross-entropy loss from candidate and non-candidate labels respectively, and dynamically differentiate the two types of OOC examples based on specially designed criteria. Then, for closed-set OOC examples, we conduct reversed label disambiguation in the non-candidate label set; for open-set OOC examples, we leverage them for training by utilizing an effective regularization strategy that dynamically assigns random candidate labels from the candidate label set. In this way, the two types of OOC examples can be differentiated and further leveraged for model training. Extensive experiments demonstrate that our proposed method outperforms state-of-the-art PLL methods.
Authors: Gavriel Habib, Noa Barzilay, Or Shimshi, Rami Ben-Ari, Nir Darshan
Gait Recognition is a computer vision task aiming to identify people by their walking patterns. Although existing methods often show high performance on specific datasets, they lack the ability to generalize to unseen scenarios. Unsupervised Domain Adaptation (UDA) tries to adapt a model, pre-trained in a supervised manner on a source domain, to an unlabelled target domain. There are only a few works on UDA for gait recognition proposing solutions to limited scenarios. In this paper, we reveal a fundamental phenomenon in adaptation of gait recognition models, caused by the bias in the target domain to viewing angle or walking direction. We then suggest a remedy to reduce this bias with a novel triplet selection strategy combined with curriculum learning. To this end, we present Gait Orientation-based method for Unsupervised Domain Adaptation (GOUDA). We provide extensive experiments on four widely-used gait datasets, CASIA-B, OU-MVLP, GREW, and Gait3D, and on three backbones, GaitSet, GaitPart, and GaitGL, justifying the view bias and showing the superiority of our proposed method over prior UDA works.
Authors: Maosu Li, Yijie Wu, Anthony G.O. Yeh, Fan Xue
Many existing 3D semantic segmentation methods, deep learning in computer vision notably, claimed to achieve desired results on urban point clouds. Thus, it is significant to assess these methods quantitatively in diversified real-world urban scenes, encompassing high-rise, low-rise, high-density, and low-density urban areas. However, existing public benchmark datasets primarily represent low-rise scenes from European cities and cannot assess the methods comprehensively. This paper presents a benchmark dataset of high-rise urban point clouds, namely High-Rise, High-Density urban scenes of Hong Kong (HRHD-HK). HRHD-HK arranged in 150 tiles contains 273 million colorful photogrammetric 3D points from diverse urban settings. The semantic labels of HRHD-HK include building, vegetation, road, waterbody, facility, terrain, and vehicle. To our best knowledge, HRHD-HK is the first photogrammetric dataset that focuses on HRHD urban areas. This paper also comprehensively evaluates eight popular semantic segmentation methods on the HRHD-HK dataset. Experimental results confirmed plenty of room for enhancing the current 3D semantic segmentation of point clouds, especially for city objects with small volumes. Our dataset is publicly available at https://doi.org/10.25442/hku.23701866.v2.
Authors: Xugui Zhou, Anqi Chen, Maxfield Kouzel, Haotian Ren, Morgan McCarty, Cristina Nita-Rotaru, Homa Alemzadeh
Adaptive Cruise Control (ACC) is a widely used driver assistance technology for maintaining the desired speed and safe distance to the leading vehicle. This paper evaluates the security of the deep neural network (DNN) based ACC systems under runtime stealthy perception attacks that strategically inject perturbations into camera data to cause forward collisions. We present a context-aware strategy for the selection of the most critical times for triggering the attacks and a novel optimization-based method for the adaptive generation of image perturbations at runtime. We evaluate the effectiveness of the proposed attack using a publicly available driving dataset, an actual vehicle, and a realistic simulation platform with the control software from a production ACC system, a physical-world driving simulator, and interventions by the human driver and safety features such as Advanced Emergency Braking System (AEBS). Experimental results show that the proposed attack achieves 142.9 times higher success rate in causing hazards and 89.6% higher evasion rate than baselines while being stealthy and robust to real-world factors and dynamic changes in the environment. This study highlights the role of human drivers and basic safety mechanisms in preventing attacks.
Authors: Dogucan Yaman, Fevziye Irem Eyiokur, Leonard Bärmann, Hazim Kemal Ekenel, Alexander Waibel
Audio-driven talking face generation is the task of creating a lip-synchronized, realistic face video from given audio and reference frames. This involves two major challenges: overall visual quality of generated images on the one hand, and audio-visual synchronization of the mouth part on the other hand. In this paper, we start by identifying several problematic aspects of synchronization methods in recent audio-driven talking face generation approaches. Specifically, this involves unintended flow of lip, pose and other information from the reference to the generated image, as well as instabilities during model training. Subsequently, we propose various techniques for obviating these issues: First, a silent-lip reference image generator prevents leaking of lips from the reference to the generated image. Second, an adaptive triplet loss handles the pose leaking problem. Finally, we propose a stabilized formulation of synchronization loss, circumventing aforementioned training instabilities while additionally further alleviating the lip leaking issue. Combining the individual improvements, we present state-of-the-art visual quality and synchronization performance on LRS2 in five out of seven and LRW in six out of seven metrics, and competitive results on the remaining ones. We further validate our design in various ablation experiments, confirming the individual contributions as well as their complementary effects.
Authors: Fu-Hsien Huang, Hsin-Min Lu
Handwritten signature verification, crucial for legal and financial institutions, faces challenges including inter-writer similarity, intra-writer variations, and limited signature samples. To address these, we introduce a MultiScale Signature feature learning Network (MS-SigNet) with a novel metric learning loss called the co-tuplet loss, designed for offline handwritten signature verification. MS-SigNet learns both global and regional signature features from multiple spatial scales, enhancing feature discrimination. This approach effectively distinguishes genuine signatures from skilled forgeries by capturing overall strokes and detailed local differences. The co-tuplet loss, focusing on multiple positive and negative examples, overcomes the limitations of typical metric learning losses by addressing inter-writer similarity and intra-writer variations and emphasizing informative examples. We also present HanSig, a large-scale Chinese signature dataset (available at https://github.com/ashleyfhh/HanSig) to support robust system development. Experimental results on four benchmark datasets in different languages demonstrate the promising performance of our method in comparison to state-of-the-art approaches.
Authors: Johan Edstedt, Georg Bökman, Mårten Wadenbäck, Michael Felsberg
Keypoint detection is a pivotal step in 3D reconstruction, whereby sets of (up to) K points are detected in each view of a scene. Crucially, the detected points need to be consistent between views, i.e., correspond to the same 3D point in the scene. One of the main challenges with keypoint detection is the formulation of the learning objective. Previous learning-based methods typically jointly learn descriptors with keypoints, and treat the keypoint detection as a binary classification task on mutual nearest neighbours. However, basing keypoint detection on descriptor nearest neighbours is a proxy task, which is not guaranteed to produce 3D-consistent keypoints. Furthermore, this ties the keypoints to a specific descriptor, complicating downstream usage. In this work, we instead learn keypoints directly from 3D consistency. To this end, we train the detector to detect tracks from large-scale SfM. As these points are often overly sparse, we derive a semi-supervised two-view detection objective to expand this set to a desired number of detections. To train a descriptor, we maximize the mutual nearest neighbour objective over the keypoints with a separate network. Results show that our approach, DeDoDe, achieves significant gains on multiple geometry benchmarks. Code is provided at https://github.com/Parskatt/DeDoDe
Authors: Heng Wang, Jianbo Ma, Santiago Pascual, Richard Cartwright, Weidong Cai
Building artificial intelligence (AI) systems on top of a set of foundation models (FMs) is becoming a new paradigm in AI research. Their representative and generative abilities learnt from vast amounts of data can be easily adapted and transferred to a wide range of downstream tasks without extra training from scratch. However, leveraging FMs in cross-modal generation remains under-researched when audio modality is involved. On the other hand, automatically generating semantically-relevant sound from visual input is an important problem in cross-modal generation studies. To solve this vision-to-audio (V2A) generation problem, existing methods tend to design and build complex systems from scratch using modestly sized datasets. In this paper, we propose a lightweight solution to this problem by leveraging foundation models, specifically CLIP, CLAP, and AudioLDM. We first investigate the domain gap between the latent space of the visual CLIP and the auditory CLAP models. Then we propose a simple yet effective mapper mechanism (V2A-Mapper) to bridge the domain gap by translating the visual input between CLIP and CLAP spaces. Conditioned on the translated CLAP embedding, pretrained audio generative FM AudioLDM is adopted to produce high-fidelity and visually-aligned sound. Compared to previous approaches, our method only requires a quick training of the V2A-Mapper. We further analyze and conduct extensive experiments on the choice of the V2A-Mapper and show that a generative mapper is better at fidelity and variability (FD) while a regression mapper is slightly better at relevance (CS). Both objective and subjective evaluation on two V2A datasets demonstrate the superiority of our proposed method compared to current state-of-the-art approaches - trained with 86% fewer parameters but achieving 53% and 19% improvement in FD and CS, respectively.
Authors: Xi Li, Songhe Wang, Ruiquan Huang, Mahanth Gowda, George Kesidis
Deep neural networks (DNNs) have achieved tremendous success in various applications including video action recognition, yet remain vulnerable to backdoor attacks (Trojans). The backdoor-compromised model will mis-classify to the target class chosen by the attacker when a test instance (from a non-target class) is embedded with a specific trigger, while maintaining high accuracy on attack-free instances. Although there are extensive studies on backdoor attacks against image data, the susceptibility of video-based systems under backdoor attacks remains largely unexplored. Current studies are direct extensions of approaches proposed for image data, e.g., the triggers are independently embedded within the frames, which tend to be detectable by existing defenses. In this paper, we introduce a simple yet effective backdoor attack against video data. Our proposed attack, adding perturbations in a transformed domain, plants an imperceptible, temporally distributed trigger across the video frames, and is shown to be resilient to existing defensive strategies. The effectiveness of the proposed attack is demonstrated by extensive experiments with various well-known models on two video recognition benchmarks, UCF101 and HMDB51, and a sign language recognition benchmark, Greek Sign Language (GSL) dataset. We delve into the impact of several influential factors on our proposed attack and identify an intriguing effect termed "collateral damage" through extensive studies.
Authors: Songchun Zhang, Chunhui Zhao
Weakly supervised temporal action localization (WSTAL) aims to localize actions in untrimmed videos using video-level labels. Despite recent advances, existing approaches mainly follow a localization-by-classification pipeline, generally processing each segment individually, thereby exploiting only limited contextual information. As a result, the model will lack a comprehensive understanding (e.g. appearance and temporal structure) of various action patterns, leading to ambiguity in classification learning and temporal localization. Our work addresses this from a novel perspective, by exploring and exploiting the cross-video contextual knowledge within the dataset to recover the dataset-level semantic structure of action instances via weak labels only, thereby indirectly improving the holistic understanding of fine-grained action patterns and alleviating the aforementioned ambiguities. Specifically, an end-to-end framework is proposed, including a Robust Memory-Guided Contrastive Learning (RMGCL) module and a Global Knowledge Summarization and Aggregation (GKSA) module. First, the RMGCL module explores the contrast and consistency of cross-video action features, assisting in learning more structured and compact embedding space, thus reducing ambiguity in classification learning. Further, the GKSA module is used to efficiently summarize and propagate the cross-video representative action knowledge in a learnable manner to promote holistic action patterns understanding, which in turn allows the generation of high-confidence pseudo-labels for self-learning, thus alleviating ambiguity in temporal localization. Extensive experiments on THUMOS14, ActivityNet1.3, and FineAction demonstrate that our method outperforms the state-of-the-art methods, and can be easily plugged into other WSTAL methods.
Authors: Qiu Zhou, Jinming Cao, Hanchao Leng, Yifang Yin, Yu Kun, Roger Zimmermann
In the field of autonomous driving, accurate and comprehensive perception of the 3D environment is crucial. Bird's Eye View (BEV) based methods have emerged as a promising solution for 3D object detection using multi-view images as input. However, existing 3D object detection methods often ignore the physical context in the environment, such as sidewalk and vegetation, resulting in sub-optimal performance. In this paper, we propose a novel approach called SOGDet (Semantic-Occupancy Guided Multi-view 3D Object Detection), that leverages a 3D semantic-occupancy branch to improve the accuracy of 3D object detection. In particular, the physical context modeled by semantic occupancy helps the detector to perceive the scenes in a more holistic view. Our SOGDet is flexible to use and can be seamlessly integrated with most existing BEV-based methods. To evaluate its effectiveness, we apply this approach to several state-of-the-art baselines and conduct extensive experiments on the exclusive nuScenes dataset. Our results show that SOGDet consistently enhance the performance of three baseline methods in terms of nuScenes Detection Score (NDS) and mean Average Precision (mAP). This indicates that the combination of 3D object detection and 3D semantic occupancy leads to a more comprehensive perception of the 3D environment, thereby aiding build more robust autonomous driving systems. The codes are available at: https://github.com/zhouqiu/SOGDet.
Authors: Jianning Li, Zongwei Zhou, Jiancheng Yang, Antonio Pepe, Christina Gsaxner, Gijs Luijten, Chongyu Qu, Tiezheng Zhang, Xiaoxi Chen, Wenxuan Li, Marek Wodzinski, Paul Friedrich, Kangxian Xie, Yuan Jin, Narmada Ambigapathy, Enrico Nasca, Naida Solak, Gian Marco Melito, Viet Duc Vu, Afaque R. Memon, Christopher Schlachta, Sandrine De Ribaupierre, Rajnikant Patel, Roy Eagleson, Xiaojun Chen, Heinrich Mächler, Jan Stefan Kirschke, Ezequiel de la Rosa, Patrick Ferdinand Christ, Hongwei Bran Li, David G. Ellis, Michele R. Aizenberg, Sergios Gatidis, Thomas Küstner, Nadya Shusharina, Nicholas Heller, Vincent Andrearczyk, Adrien Depeursinge, Mathieu Hatt, Anjany Sekuboyina, Maximilian Löffler, Hans Liebl, Reuben Dorent, Tom Vercauteren, Jonathan Shapey, Aaron Kujawa, Stefan Cornelissen, et al. (110 additional authors not shown)
Prior to the deep learning era, \textit{shape} was commonly used to describe the objects. Nowadays, state-of-the-art (SOTA) algorithms in medical imaging are predominantly diverging from computer vision, where voxel grids, meshes, point clouds, and implicit surface models are used. This is seen from numerous shape-related publications in premier vision conferences as well as the growing popularity of \textit{ShapeNet} (about 51,300 models) and \textit{Princeton ModelNet} (127,915 models). For the medical domain, we present a large collection of anatomical shapes (e.g., bones, organs, vessels) and 3D models of surgical instrument, called \textit{MedShapeNet}, created to facilitate the translation of data-driven vision algorithms to medical applications and to adapt SOTA vision algorithms to medical problems. As a unique feature, we directly model the majority of shapes on the imaging data of real patients. As of today, \textit{MedShapeNet} includes 23 dataset with more than 100,000 shapes that are paired with annotations (ground truth). Our data is freely accessible via a web interface and a Python application programming interface (API) and can be used for discriminative, reconstructive, and variational benchmarks as well as various applications in virtual, augmented, or mixed reality, and 3D printing. Exemplary, we present use cases in the fields of classification of brain tumors, facial and skull reconstructions, multi-class anatomy completion, education, and 3D printing. In future, we will extend the data and improve the interfaces. The project pages are: \url{https://medshapenet.ikim.nrw/} and \url{https://github.com/Jianningli/medshapenet-feedback}
Authors: Yuanshuo Cheng, Mingwen Shao, Yecong Wan, Yuanjian Qiao, Wangmeng Zuo, Deyu Meng
Existing Video Restoration (VR) methods always necessitate the individual deployment of models for each adverse weather to remove diverse adverse weather degradations, lacking the capability for adaptive processing of degradations. Such limitation amplifies the complexity and deployment costs in practical applications. To overcome this deficiency, in this paper, we propose a Cross-consistent Deep Unfolding Network (CDUN) for All-In-One VR, which enables the employment of a single model to remove diverse degradations for the first time. Specifically, the proposed CDUN accomplishes a novel iterative optimization framework, capable of restoring frames corrupted by corresponding degradations according to the degradation features given in advance. To empower the framework for eliminating diverse degradations, we devise a Sequence-wise Adaptive Degradation Estimator (SADE) to estimate degradation features for the input corrupted video. By orchestrating these two cascading procedures, CDUN achieves adaptive processing for diverse degradation. In addition, we introduce a window-based inter-frame fusion strategy to utilize information from more adjacent frames. This strategy involves the progressive stacking of temporal windows in multiple iterations, effectively enlarging the temporal receptive field and enabling each frame's restoration to leverage information from distant frames. Extensive experiments demonstrate that the proposed method achieves state-of-the-art performance in All-In-One VR.
Authors: Alhasan Abdellatif, Ahmed H. Elsheikh
In this paper, we introduce a novel approach for generating texture images of infinite sizes using Generative Adversarial Networks (GANs) based on a patch-by-patch paradigm. Existing texture synthesis techniques rely on generating large-scale textures using a single forward pass to the generative model; this approach limits the scalability and flexibility of the images produced. In contrast, the proposed approach trains a GAN model on a single texture image to generate relatively small-size patches that are locally correlated and can be seamlessly concatenated to form a larger image. The method relies on local padding in the generator to ensure consistency between the generated patches. It also utilizes spatial stochastic modulation to allow for local variations and improve patterns alignment in the large-scale image. The trained models learn the local texture structure and are able to generate images of arbitrary sizes, while also maintaining the coherence and diversity. Experimental results demonstrate constant GPU scalability with respect to the generated image size compared to existing approaches that exhibit a proportional growth in GPU memory.
Authors: Sergi Garcia-Bordils, Dimosthenis Karatzas, Marçal Rusiñol
We introduce the structured scene-text spotting task, which requires a scene-text OCR system to spot text in the wild according to a query regular expression. Contrary to generic scene text OCR, structured scene-text spotting seeks to dynamically condition both scene text detection and recognition on user-provided regular expressions. To tackle this task, we propose the Structured TExt sPotter (STEP), a model that exploits the provided text structure to guide the OCR process. STEP is able to deal with regular expressions that contain spaces and it is not bound to detection at the word-level granularity. Our approach enables accurate zero-shot structured text spotting in a wide variety of real-world reading scenarios and is solely trained on publicly available data. To demonstrate the effectiveness of our approach, we introduce a new challenging test dataset that contains several types of out-of-vocabulary structured text, reflecting important reading applications of fields such as prices, dates, serial numbers, license plates etc. We demonstrate that STEP can provide specialised OCR performance on demand in all tested scenarios.
Authors: Shaheer Mohamed, Maryam Haghighat, Tharindu Fernando, Sridha Sridharan, Clinton Fookes, Peyman Moghadam
Hyperspectral images (HSIs) contain rich spectral and spatial information. Motivated by the success of transformers in the field of natural language processing and computer vision where they have shown the ability to learn long range dependencies within input data, recent research has focused on using transformers for HSIs. However, current state-of-the-art hyperspectral transformers only tokenize the input HSI sample along the spectral dimension, resulting in the under-utilization of spatial information. Moreover, transformers are known to be data-hungry and their performance relies heavily on large-scale pre-training, which is challenging due to limited annotated hyperspectral data. Therefore, the full potential of HSI transformers has not been fully realized. To overcome these limitations, we propose a novel factorized spectral-spatial transformer that incorporates factorized self-supervised pre-training procedures, leading to significant improvements in performance. The factorization of the inputs allows the spectral and spatial transformers to better capture the interactions within the hyperspectral data cubes. Inspired by masked image modeling pre-training, we also devise efficient masking strategies for pre-training each of the spectral and spatial transformers. We conduct experiments on six publicly available datasets for HSI classification task and demonstrate that our model achieves state-of-the-art performance in all the datasets. The code for our model will be made available at https://github.com/csiro-robotics/factoformer.
Authors: Tao Pu, Tianshui Chen, Hefeng Wu, Yongyi Lu, Liang Lin
Video scene graph generation (VidSGG) aims to identify objects in visual scenes and infer their relationships for a given video. It requires not only a comprehensive understanding of each object scattered on the whole scene but also a deep dive into their temporal motions and interactions. Inherently, object pairs and their relationships enjoy spatial co-occurrence correlations within each image and temporal consistency/transition correlations across different images, which can serve as prior knowledge to facilitate VidSGG model learning and inference. In this work, we propose a spatial-temporal knowledge-embedded transformer (STKET) that incorporates the prior spatial-temporal knowledge into the multi-head cross-attention mechanism to learn more representative relationship representations. Specifically, we first learn spatial co-occurrence and temporal transition correlations in a statistical manner. Then, we design spatial and temporal knowledge-embedded layers that introduce the multi-head cross-attention mechanism to fully explore the interaction between visual representation and the knowledge to generate spatial- and temporal-embedded representations, respectively. Finally, we aggregate these representations for each subject-object pair to predict the final semantic labels and their relationships. Extensive experiments show that STKET outperforms current competing algorithms by a large margin, e.g., improving the mR@50 by 8.1%, 4.7%, and 2.1% on different settings over current algorithms.
Authors: Trong Thang Pham, Jacob Brecheisen, Anh Nguyen, Hien Nguyen, Ngan Le
In the field of chest X-ray (CXR) diagnosis, existing works often focus solely on determining where a radiologist looks, typically through tasks such as detection, segmentation, or classification. However, these approaches are often designed as black-box models, lacking interpretability. In this paper, we introduce Interpretable Artificial Intelligence (I-AI) a novel and unified controllable interpretable pipeline for decoding the intense focus of radiologists in CXR diagnosis. Our I-AI addresses three key questions: where a radiologist looks, how long they focus on specific areas, and what findings they diagnose. By capturing the intensity of the radiologist's gaze, we provide a unified solution that offers insights into the cognitive process underlying radiological interpretation. Unlike current methods that rely on black-box machine learning models, which can be prone to extracting erroneous information from the entire input image during the diagnosis process, we tackle this issue by effectively masking out irrelevant information. Our proposed I-AI leverages a vision-language model, allowing for precise control over the interpretation process while ensuring the exclusion of irrelevant features. To train our I-AI model, we utilize an eye gaze dataset to extract anatomical gaze information and generate ground truth heatmaps. Through extensive experimentation, we demonstrate the efficacy of our method. We showcase that the attention heatmaps, designed to mimic radiologists' focus, encode sufficient and relevant information, enabling accurate classification tasks using only a portion of CXR. The code, checkpoints, and data are at https://github.com/UARK-AICV/IAI
Authors: Saeejith Nair, Yuhao Chen, Mohammad Javad Shafiee, Alexander Wong
Neural radiance fields (NeRFs) enable high-quality novel view synthesis, but their high computational complexity limits deployability. While existing neural-based solutions strive for efficiency, they use one-size-fits-all architectures regardless of scene complexity. The same architecture may be unnecessarily large for simple scenes but insufficient for complex ones. Thus, there is a need to dynamically optimize the neural network component of NeRFs to achieve a balance between computational complexity and specific targets for synthesis quality. We introduce NAS-NeRF, a generative neural architecture search strategy that generates compact, scene-specialized NeRF architectures by balancing architecture complexity and target synthesis quality metrics. Our method incorporates constraints on target metrics and budgets to guide the search towards architectures tailored for each scene. Experiments on the Blender synthetic dataset show the proposed NAS-NeRF can generate architectures up to 5.74$\times$ smaller, with 4.19$\times$ fewer FLOPs, and 1.93$\times$ faster on a GPU than baseline NeRFs, without suffering a drop in SSIM. Furthermore, we illustrate that NAS-NeRF can also achieve architectures up to 23$\times$ smaller, with 22$\times$ fewer FLOPs, and 4.7$\times$ faster than baseline NeRFs with only a 5.3% average SSIM drop. Our source code is also made publicly available at https://saeejithnair.github.io/NAS-NeRF.
Authors: Xin Yuan, Michael Maire
We develop a neural network architecture which, trained in an unsupervised manner as a denoising diffusion model, simultaneously learns to both generate and segment images. Learning is driven entirely by the denoising diffusion objective, without any annotation or prior knowledge about regions during training. A computational bottleneck, built into the neural architecture, encourages the denoising network to partition an input into regions, denoise them in parallel, and combine the results. Our trained model generates both synthetic images and, by simple examination of its internal predicted partitions, a semantic segmentation of those images. Without any finetuning, we directly apply our unsupervised model to the downstream task of segmenting real images via noising and subsequently denoising them. Experiments demonstrate that our model achieves accurate unsupervised image segmentation and high-quality synthetic image generation across multiple datasets.
Authors: Hao Zhang, Lumin Xu, Shenqi Lai, Wenqi Shao, Nanning Zheng, Ping Luo, Yu Qiao, Kaipeng Zhang
Current image-based keypoint detection methods for animal (including human) bodies and faces are generally divided into full-supervised and few-shot class-agnostic approaches. The former typically relies on laborious and time-consuming manual annotations, posing considerable challenges in expanding keypoint detection to a broader range of keypoint categories and animal species. The latter, though less dependent on extensive manual input, still requires necessary support images with annotation for reference during testing. To realize zero-shot keypoint detection without any prior annotation, we introduce the Open-Vocabulary Keypoint Detection (OVKD) task, which is innovatively designed to use text prompts for identifying arbitrary keypoints across any species. In pursuit of this goal, we have developed a novel framework named Open-Vocabulary Keypoint Detection with Semantic-feature Matching (KDSM). This framework synergistically combines vision and language models, creating an interplay between language features and local keypoint visual features. KDSM enhances its capabilities by integrating Domain Distribution Matrix Matching (DDMM) and other special modules, such as the Vision-Keypoint Relational Awareness (VKRA) module, improving the framework's generalizability and overall performance.Our comprehensive experiments demonstrate that KDSM significantly outperforms the baseline in terms of performance and achieves remarkable success in the OVKD task.Impressively, our method, operating in a zero-shot fashion, still yields results comparable to state-of-the-art few-shot species class-agnostic keypoint detection methods.We will make the source code publicly accessible.
Authors: Rong Wang, Wei Mao, Hongdong Li
This paper addresses the task of 3D pose estimation for a hand interacting with an object from a single image observation. When modeling hand-object interaction, previous works mainly exploit proximity cues, while overlooking the dynamical nature that the hand must stably grasp the object to counteract gravity and thus preventing the object from slipping or falling. These works fail to leverage dynamical constraints in the estimation and consequently often produce unstable results. Meanwhile, refining unstable configurations with physics-based reasoning remains challenging, both by the complexity of contact dynamics and by the lack of effective and efficient physics inference in the data-driven learning framework. To address both issues, we present DeepSimHO: a novel deep-learning pipeline that combines forward physics simulation and backward gradient approximation with a neural network. Specifically, for an initial hand-object pose estimated by a base network, we forward it to a physics simulator to evaluate its stability. However, due to non-smooth contact geometry and penetration, existing differentiable simulators can not provide reliable state gradient. To remedy this, we further introduce a deep network to learn the stability evaluation process from the simulator, while smoothly approximating its gradient and thus enabling effective back-propagation. Extensive experiments show that our method noticeably improves the stability of the estimation and achieves superior efficiency over test-time optimization. The code is available at https://github.com/rongakowang/DeepSimHO.
Authors: Ancheng Lin, Jun Li
Neural Radiance Fields (NeRFs) have shown great potential in modelling 3D scenes. Dynamic NeRFs extend this model by capturing time-varying elements, typically using deformation fields. The existing dynamic NeRFs employ a similar Eulerian representation for both light radiance and deformation fields. This leads to a close coupling of appearance and motion and lacks a physical interpretation. In this work, we propose Dynamic Appearance Particle Neural Radiance Field (DAP-NeRF), which introduces particle-based representation to model the motions of visual elements in a dynamic 3D scene. DAP-NeRF consists of superposition of a static field and a dynamic field. The dynamic field is quantised as a collection of {\em appearance particles}, which carries the visual information of a small dynamic element in the scene and is equipped with a motion model. All components, including the static field, the visual features and motion models of the particles, are learned from monocular videos without any prior geometric knowledge of the scene. We develop an efficient computational framework for the particle-based model. We also construct a new dataset to evaluate motion modelling. Experimental results show that DAP-NeRF is an effective technique to capture not only the appearance but also the physically meaningful motions in a 3D dynamic scene.
Authors: Roy Kapon, Guy Tevet, Daniel Cohen-Or, Amit H. Bermano
We introduce Multi-view Ancestral Sampling (MAS), a method for 3D motion generation, using 2D diffusion models that were trained on motions obtained from in-the-wild videos. As such, MAS opens opportunities to exciting and diverse fields of motion previously under-explored as 3D data is scarce and hard to collect. MAS works by simultaneously denoising multiple 2D motion sequences representing different views of the same 3D motion. It ensures consistency across all views at each diffusion step by combining the individual generations into a unified 3D sequence, and projecting it back to the original views. We demonstrate MAS on 2D pose data acquired from videos depicting professional basketball maneuvers, rhythmic gymnastic performances featuring a ball apparatus, and horse races. In each of these domains, 3D motion capture is arduous, and yet, MAS generates diverse and realistic 3D sequences. Unlike the Score Distillation approach, which optimizes each sample by repeatedly applying small fixes, our method uses a sampling process that was constructed for the diffusion framework. As we demonstrate, MAS avoids common issues such as out-of-domain sampling and mode-collapse. https://guytevet.github.io/mas-page/
Authors: Xin Yuan, Jie Guo, Weidong Qiu, Zheng Huang, Shujun Li
Mis- and disinformation online have become a major societal problem as major sources of online harms of different kinds. One common form of mis- and disinformation is out-of-context (OOC) information, where different pieces of information are falsely associated, e.g., a real image combined with a false textual caption or a misleading textual description. Although some past studies have attempted to defend against OOC mis- and disinformation through external evidence, they tend to disregard the role of different pieces of evidence with different stances. Motivated by the intuition that the stance of evidence represents a bias towards different detection results, we propose a stance extraction network (SEN) that can extract the stances of different pieces of multi-modal evidence in a unified framework. Moreover, we introduce a support-refutation score calculated based on the co-occurrence relations of named entities into the textual SEN. Extensive experiments on a public large-scale dataset demonstrated that our proposed method outperformed the state-of-the-art baselines, with the best model achieving a performance gain of 3.2% in accuracy.
Authors: Dan Song, Xuanpu Zhang, Juan Zhou, Weizhi Nie, Ruofeng Tong, Mohan Kankanhalli, An-An Liu
Image-based virtual try-on aims to synthesize a naturally dressed person image with a clothing image, which revolutionizes online shopping and inspires related topics within image generation, showing both research significance and commercial potential. However, there is a big gap between current research progress and commercial applications and an absence of comprehensive overview of this field to accelerate the development. In this survey, we provide a comprehensive analysis of the state-of-the-art techniques and methodologies in aspects of pipeline architecture, person representation and key modules such as try-on indication, clothing warping and try-on stage. We propose a new semantic criteria with CLIP, and evaluate representative methods with uniformly implemented evaluation metrics on the same dataset. In addition to quantitative and qualitative evaluation of current open-source methods, we also utilize ControlNet to fine-tune a recent large image generation model (PBE) to show future potential of large-scale models on image-based virtual try-on task. Finally, unresolved issues are highlighted and future research directions are prospected to identify key trends and inspire further exploration. The uniformly implemented evaluation metrics, dataset and collected methods will be made public available at https://github.com/little-misfit/Survey-Of-Virtual-Try-On.
Authors: Fei Wu, Pablo Marquez-Neila, Mingyi Zheng, Hedyeh Rafii-Tari, Raphael Sznitman
Semantic segmentation is a complex task that relies heavily on large amounts of annotated image data. However, annotating such data can be time-consuming and resource-intensive, especially in the medical domain. Active Learning (AL) is a popular approach that can help to reduce this burden by iteratively selecting images for annotation to improve the model performance. In the case of video data, it is important to consider the model uncertainty and the temporal nature of the sequences when selecting images for annotation. This work proposes a novel AL strategy for surgery video segmentation, COWAL, COrrelation-aWare Active Learning. Our approach involves projecting images into a latent space that has been fine-tuned using contrastive learning and then selecting a fixed number of representative images from local clusters of video frames. We demonstrate the effectiveness of this approach on two video datasets of surgical instruments and three real-world video datasets. The datasets and code will be made publicly available upon receiving necessary approvals.
Authors: Marwah Sulaiman, Zahraa Shehabeldin, Israa Fahmy, Mohammed Barakat, Mohammed El-Naggar, Dareen Hussein, Moustafa Youssef, Hesham M. Eraqi
Recently, video super resolution (VSR) has become a very impactful task in the area of Computer Vision due to its various applications. In this paper, we propose Recurrent Back-Projection Generative Adversarial Network (RBPGAN) for VSR in an attempt to generate temporally coherent solutions while preserving spatial details. RBPGAN integrates two state-of-the-art models to get the best in both worlds without compromising the accuracy of produced video. The generator of the model is inspired by RBPN system, while the discriminator is inspired by TecoGAN. We also utilize Ping-Pong loss to increase temporal consistency over time. Our contribution together results in a model that outperforms earlier work in terms of temporally consistent details, as we will demonstrate qualitatively and quantitatively using different datasets.
Authors: Ali Borji
Although extensive research has been carried out to evaluate the effectiveness of AI tools and models in detecting deep fakes, the question remains unanswered regarding whether these models can accurately identify genuine images that appear artificial. In this study, as an initial step towards addressing this issue, we have curated a dataset of 510 genuine images that exhibit a fake appearance and conducted an assessment using two AI models. We show that two models exhibited subpar performance when applied to our dataset. Additionally, our dataset can serve as a valuable tool for assessing the ability of deep learning models to comprehend complex visual stimuli. We anticipate that this research will stimulate further discussions and investigations in this area. Our dataset is accessible at https://github.com/aliborji/FLORIDA.
Authors: Nan Zhong, Yiran Xu, Zhenxing Qian, Xinpeng Zhang
Recent generative models show impressive performance in generating photographic images. Humans can hardly distinguish such incredibly realistic-looking AI-generated images from real ones. AI-generated images may lead to ubiquitous disinformation dissemination. Therefore, it is of utmost urgency to develop a detector to identify AI-generated images. Most existing detectors suffer from sharp performance drops over unseen generative models. In this paper, we propose a novel AI-generated image detector capable of identifying fake images created by a wide range of generative models. Our approach leverages the inter-pixel correlation contrast between rich and poor texture regions within an image. Pixels in rich texture regions exhibit more significant fluctuations than those in poor texture regions. This discrepancy reflects that the entropy of rich texture regions is larger than that of poor ones. Consequently, synthesizing realistic rich texture regions proves to be more challenging for existing generative models. Based on this principle, we divide an image into multiple patches and reconstruct them into two images, comprising rich-texture and poor-texture patches respectively. Subsequently, we extract the inter-pixel correlation discrepancy feature between rich and poor texture regions. This feature serves as a universal fingerprint used for AI-generated image forensics across different generative models. In addition, we build a comprehensive AI-generated image detection benchmark, which includes 16 kinds of prevalent generative models, to evaluate the effectiveness of existing baselines and our approach. Our benchmark provides a leaderboard for follow-up studies. Extensive experimental results show that our approach outperforms state-of-the-art baselines by a significant margin. Our project: https://fdmas.github.io/AIGCDetect/
Authors: Yuheng Xue, Nenglun Chen, Jun Liu, Wenyun Sun
Recently, many 2D pretrained foundational models have demonstrated impressive zero-shot prediction capabilities. In this work, we design a novel pipeline for zero-shot 3D part segmentation, called ZeroPS. It high-quality transfers knowledge from 2D pretrained foundational models to 3D point clouds. The main idea of our approach is to explore the natural relationship between multi-view correspondences and the prompt mechanism of foundational models and build bridges on it. Our pipeline consists of two components: 1) a self-extension component that extends 2D groups from a single viewpoint to spatial global-level 3D groups; 2) a multi-modal labeling component that introduces a two-dimensional checking mechanism to vote each 2D predicted bounding box to the best matching 3D part, and a Class Non-highest Vote Penalty function to refine the Vote Matrix. Additionally, a merging algorithm is included to merge part-level 3D groups. Extensive evaluation of three zero-shot segmentation tasks on PartnetE datasets, achieving state-of-the-art results with significant improvements (+19.6%, +5.2% and +4.9%, respectively) over existing methods. Our proposed approach does not need any training, fine-tuning or learnable parameters. It is hardly affected by domain shift. The code will be released.
Authors: Kai Cheng, Xiaoxiao Long, Wei Yin, Jin Wang, Zhiqiang Wu, Yuexin Ma, Kaixuan Wang, Xiaozhi Chen, Xuejin Chen
Multi-camera setups find widespread use across various applications, such as autonomous driving, as they greatly expand sensing capabilities. Despite the fast development of Neural radiance field (NeRF) techniques and their wide applications in both indoor and outdoor scenes, applying NeRF to multi-camera systems remains very challenging. This is primarily due to the inherent under-calibration issues in multi-camera setup, including inconsistent imaging effects stemming from separately calibrated image signal processing units in diverse cameras, and system errors arising from mechanical vibrations during driving that affect relative camera poses. In this paper, we present UC-NeRF, a novel method tailored for novel view synthesis in under-calibrated multi-view camera systems. Firstly, we propose a layer-based color correction to rectify the color inconsistency in different image regions. Second, we propose virtual warping to generate more viewpoint-diverse but color-consistent virtual views for color correction and 3D recovery. Finally, a spatiotemporally constrained pose refinement is designed for more robust and accurate pose calibration in multi-camera systems. Our method not only achieves state-of-the-art performance of novel view synthesis in multi-camera setups, but also effectively facilitates depth estimation in large-scale outdoor scenes with the synthesized novel views.
Authors: Fangxin Shang, Jie Fu, Yehui Yang, Haifeng Huang, Junwei Liu, Lei Ma
In the field of medical imaging, there are seldom large-scale public datasets with high-quality annotations due to data privacy and annotation cost. To address this issue, we release SynFundus-1M, a high-quality synthetic dataset containing over \textbf{1 million} fundus images w.r.t. 11 disease types. Moreover, we intentionally diversify the readability of the images and accordingly provide 4 types of the quality score for each image. To the best of our knowledge, SynFundus-1M is currently the largest fundus dataset with the most sophisticated annotations. All the images are generated by a Denoising Diffusion Probabilistic Model, named SynFundus-Generator. Trained with over 1.3 million private fundus images, our SynFundus-Generator achieves significant superior performance in generating fundus images compared to some recent related works. Furthermore, we blend some synthetic images from SynFundus-1M with real fundus images, and ophthalmologists can hardly distinguish the synthetic images from real ones. Through extensive experiments, we demonstrate that both convolutional neural networs (CNN) and Vision Transformer (ViT) can benefit from SynFundus-1M by pretraining or training directly. Compared to datasets like ImageNet or EyePACS, models trained on SynFundus-1M not only achieve better performance but also faster convergence on various downstream tasks.
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: Qiang Wen, Yazhou Xing, Qifeng Chen
Enhancing a low-light noisy RAW image into a well-exposed and clean sRGB image is a significant challenge in computational photography. Due to the limitation of large-scale paired data, prior approaches have difficulty in recovering fine details and true colors in extremely low-light regions. Meanwhile, recent advancements in generative diffusion models have shown promising generating capabilities, which inspires this work to explore generative priors from a diffusion model trained on a large-scale open-domain dataset to benefit the low-light image enhancement (LLIE) task. Based on this intention, we propose a novel diffusion-model-based LLIE method, dubbed LDM-SID. LDM-SID aims at inserting a set of proposed taming modules into a frozen pre-trained diffusion model to steer its generating process. Specifically, the taming module fed with low-light information serves to output a pair of affine transformation parameters to modulate the intermediate feature in the diffusion model. Additionally, based on the observation of dedicated generative priors across different portions of the diffusion model, we propose to apply 2D discrete wavelet transforms on the input RAW image, resulting in dividing the LLIE task into two essential parts: low-frequency content generation and high-frequency detail maintenance. This enables us to skillfully tame the diffusion model for optimized structural generation and detail enhancement. Extensive experiments demonstrate the proposed method not only achieves state-of-the-art performance in quantitative evaluations but also shows significant superiority in visual comparisons. These findings highlight the effectiveness of leveraging a pre-trained diffusion model as a generative prior to the LLIE task. The project page is available at https://csqiangwen.github.io/projects/ldm-sid/
Authors: Kaiwen Yang, Tao Shen, Xinmei Tian, Xiubo Geng, Chongyang Tao, Dacheng Tao, Tianyi Zhou
Aligning the recent large language models (LLMs) with computer vision models leads to large vision-language models (LVLMs), which have paved the way for zero-shot image reasoning tasks. However, LVLMs are usually trained on short high-level captions only referring to sparse focus regions in images. Such a ``tunnel vision'' limits LVLMs to exploring other relevant contexts in complex scenes. To address this challenge, we introduce Question-Driven Visual Exploration (QVix), a novel prompting strategy that enhances the exploratory capabilities of LVLMs in zero-shot reasoning tasks. QVix leverages LLMs' strong language prior to generate input-exploratory questions with more details than the original query, guiding LVLMs to explore visual content more comprehensively and uncover subtle or peripheral details. QVix enables a wider exploration of visual scenes, improving the LVLMs' reasoning accuracy and depth in tasks such as visual question answering and visual entailment. Our evaluations on various challenging zero-shot vision-language benchmarks, including ScienceQA and fine-grained visual classification, demonstrate that QVix significantly outperforms existing methods, highlighting its effectiveness in bridging the gap between complex visual data and LVLMs' exploratory abilities.
Authors: Mingyue Guo, Li Yuan, Zhaoyi Yan, Binghui Chen, Yaowei Wang, Qixiang Ye
Crowd counting has achieved significant progress by training regressors to predict instance positions. In heavily crowded scenarios, however, regressors are challenged by uncontrollable annotation variance, which causes density map bias and context information inaccuracy. In this study, we propose mutual prompt learning (mPrompt), which leverages a regressor and a segmenter as guidance for each other, solving bias and inaccuracy caused by annotation variance while distinguishing foreground from background. In specific, mPrompt leverages point annotations to tune the segmenter and predict pseudo head masks in a way of point prompt learning. It then uses the predicted segmentation masks, which serve as spatial constraint, to rectify biased point annotations as context prompt learning. mPrompt defines a way of mutual information maximization from prompt learning, mitigating the impact of annotation variance while improving model accuracy. Experiments show that mPrompt significantly reduces the Mean Average Error (MAE), demonstrating the potential to be general framework for down-stream vision tasks.
Authors: Ying Yuan, Haichuan Che, Yuzhe Qin, Binghao Huang, Zhao-Heng Yin, Kang-Won Lee, Yi Wu, Soo-Chul Lim, Xiaolong Wang
Executing contact-rich manipulation tasks necessitates the fusion of tactile and visual feedback. However, the distinct nature of these modalities poses significant challenges. In this paper, we introduce a system that leverages visual and tactile sensory inputs to enable dexterous in-hand manipulation. Specifically, we propose Robot Synesthesia, a novel point cloud-based tactile representation inspired by human tactile-visual synesthesia. This approach allows for the simultaneous and seamless integration of both sensory inputs, offering richer spatial information and facilitating better reasoning about robot actions. The method, trained in a simulated environment and then deployed to a real robot, is applicable to various in-hand object rotation tasks. Comprehensive ablations are performed on how the integration of vision and touch can improve reinforcement learning and Sim2Real performance. Our project page is available at https://yingyuan0414.github.io/visuotactile/ .
Authors: Christoph Hümmer, Manuel Schwonberg, Liangwei Zhou, Hu Cao, Alois Knoll, Hanno Gottschalk
Domain generalization (DG) remains a significant challenge for perception based on deep neural networks (DNN), where domain shifts occur due to lighting, weather, or geolocation changes. In this work, we propose VLTSeg to enhance domain generalization in semantic segmentation, where the network is solely trained on the source domain and evaluated on unseen target domains. Our method leverages the inherent semantic robustness of vision-language models. First, by substituting traditional vision-only backbones with pre-trained encoders from CLIP and EVA-CLIP as transfer learning setting we find that in the field of DG, vision-language pre-training significantly outperforms supervised and self-supervised vision pre-training. We thus propose a new vision-language approach for domain generalized segmentation, which improves the domain generalization SOTA by 7.6% mIoU when training on the synthetic GTA5 dataset. We further show the superior generalization capabilities of vision-language segmentation models by reaching 76.48% mIoU on the popular Cityscapes-to-ACDC benchmark, outperforming the previous SOTA approach by 6.9% mIoU on the test set at the time of writing. Additionally, our approach shows strong in-domain generalization capabilities indicated by 86.1% mIoU on the Cityscapes test set, resulting in a shared first place with the previous SOTA on the current leaderboard at the time of submission.
Authors: Xiangru Zhu, Penglei Sun, Chengyu Wang, Jingping Liu, Zhixu Li, Yanghua Xiao, Jun Huang
Text-to-image (T2I) synthesis has recently achieved significant advancements. However, challenges remain in the model's compositionality, which is the ability to create new combinations from known components. We introduce Winoground-T2I, a benchmark designed to evaluate the compositionality of T2I models. This benchmark includes 11K complex, high-quality contrastive sentence pairs spanning 20 categories. These contrastive sentence pairs with subtle differences enable fine-grained evaluations of T2I synthesis models. Additionally, to address the inconsistency across different metrics, we propose a strategy that evaluates the reliability of various metrics by using comparative sentence pairs. We use Winoground-T2I with a dual objective: to evaluate the performance of T2I models and the metrics used for their evaluation. Finally, we provide insights into the strengths and weaknesses of these metrics and the capabilities of current T2I models in tackling challenges across a range of complex compositional categories. Our benchmark is publicly available at https://github.com/zhuxiangru/Winoground-T2I .
Authors: Jamie Santos, Holly Dinkel, Julia Di, Paulo V.K. Borges, Marina Moreira, Oleg Alexandrov, Brian Coltin, Trey Smith
This work presents an algorithm for scene change detection from point clouds to enable autonomous robotic caretaking in future space habitats. Autonomous robotic systems will help maintain future deep-space habitats, such as the Gateway space station, which will be uncrewed for extended periods. Existing scene analysis software used on the International Space Station (ISS) relies on manually-labeled images for detecting changes. In contrast, the algorithm presented in this work uses raw, unlabeled point clouds as inputs. The algorithm first applies modified Expectation-Maximization Gaussian Mixture Model (GMM) clustering to two input point clouds. It then performs change detection by comparing the GMMs using the Earth Mover's Distance. The algorithm is validated quantitatively and qualitatively using a test dataset collected by an Astrobee robot in the NASA Ames Granite Lab comprising single frame depth images taken directly by Astrobee and full-scene reconstructed maps built with RGB-D and pose data from Astrobee. The runtimes of the approach are also analyzed in depth. The source code is publicly released to promote further development.
Authors: Chenhuan Li, Meihua Xiao, zehuan li, Fangping Chen, Shanshan Qiao, Dingli Wang, Mengxi Gao, Siyi Zhang
Recently, vision transformers have performed well in various computer vision tasks, including voxel 3D reconstruction. However, the windows of the vision transformer are not multi-scale, and there is no connection between the windows, which limits the accuracy of voxel 3D reconstruction. Therefore, we propose a voxel 3D reconstruction network based on shifted window attention. To the best of our knowledge, this is the first work to apply shifted window attention to voxel 3D reconstruction. Experimental results on ShapeNet verify our method achieves SOTA accuracy in single-view reconstruction.
Authors: Hao-Chun Yang, Ole Andreassen, Lars Tjelta Westlye, Andre F. Marquand, Christian F. Beckmann, Thomas Wolfers
The detection of heterogeneous mental disorders based on brain readouts remains challenging due to the complexity of symptoms and the absence of reliable biomarkers. This paper introduces CAM (Cortical Anomaly Detection through Masked Image Modeling), a novel self-supervised framework designed for the unsupervised detection of complex brain disorders using cortical surface features. We employ this framework for the detection of individuals on the psychotic spectrum and demonstrate its capabilities compared to state-ofthe-art methods, achieving an AUC of 0.696 for Schizoaffective and 0.769 for Schizophreniform, without the need for any labels. Furthermore, the analysis of atypical cortical regions includes Pars Triangularis and several frontal areas, often implicated in schizophrenia, provide further confidence in our approach. Altogether, we demonstrate a scalable approach for anomaly detection of complex brain disorders based on cortical abnormalities.
Authors: Akshat Jindal, Shreya Singh, Soham Gadgil
In this paper, we analyze different methods to mitigate inherent geographical biases present in state of the art image classification models. We first quantitatively present this bias in two datasets - The Dollar Street Dataset and ImageNet, using images with location information. We then present different methods which can be employed to reduce this bias. Finally, we analyze the effectiveness of the different techniques on making these models more robust to geographical locations of the images.
Authors: Cong Wang, Jiaxi Gu, Panwen Hu, Songcen Xu, Hang Xu, Xiaodan Liang
Image-to-video generation, which aims to generate a video starting from a given reference image, has drawn great attention. Existing methods try to extend pre-trained text-guided image diffusion models to image-guided video generation models. Nevertheless, these methods often result in either low fidelity or flickering over time due to their limitation to shallow image guidance and poor temporal consistency. To tackle these problems, we propose a high-fidelity image-to-video generation method by devising a frame retention branch on the basis of a pre-trained video diffusion model, named DreamVideo. Instead of integrating the reference image into the diffusion process in a semantic level, our DreamVideo perceives the reference image via convolution layers and concatenate the features with the noisy latents as model input. By this means, the details of the reference image can be preserved to the greatest extent. In addition, by incorporating double-condition classifier-free guidance, a single image can be directed to videos of different actions by providing varying prompt texts. This has significant implications for controllable video generation and holds broad application prospects. We conduct comprehensive experiments on the public dataset, both quantitative and qualitative results indicate that our method outperforms the state-of-the-art method. Especially for fidelity, our model has powerful image retention ability and result in high FVD in UCF101 compared to other image-to-video models. Also, precise control can be achieved by giving different text prompts. Further details and comprehensive results of our model will be presented in https://anonymous0769.github.io/DreamVideo/.
Authors: Yuexing Han, Guanxin Wan, Bing Wang
Deep learning has yielded remarkable outcomes in various domains. However, the challenge of requiring large-scale labeled samples still persists in deep learning. Thus, data augmentation has been introduced as a critical strategy to train deep learning models. However, data augmentation suffers from information loss and poor performance in small sample environments. To overcome these drawbacks, we propose a feature augmentation method based on shape space theory, i.e., Geodesic curve feature augmentation, called GCFA in brevity.First, we extract features from the image with the neural network model. Then, the multiple image features are projected into a pre-shape space as features. In the pre-shape space, a Geodesic curve is built to fit the features. Finally, the many generated features on the Geodesic curve are used to train the various machine learning models. The GCFA module can be seamlessly integrated with most machine learning methods. And the proposed method is simple, effective and insensitive for the small sample datasets.Several examples demonstrate that the GCFA method can greatly improve the performance of the data preprocessing model in a small sample environment.
Authors: Zhuoyan Liu, Bo Wang, Lizhi Wang, Chenyu Mao, Ye Li
Multimodal semantic segmentation is developing rapidly, but the modality of RGB-Polarization remains underexplored. To delve into this problem, we construct a UPLight RGB-P segmentation benchmark with 12 typical underwater semantic classes. In this work, we design the ShareCMP, an RGB-P semantic segmentation framework with a shared dual-branch architecture, which reduces the number of parameters by about 26-33% compared to previous dual-branch models. It encompasses a Polarization Generate Attention (PGA) module designed to generate polarization modal images with richer polarization properties for the encoder. In addition, we introduce the Class Polarization-Aware Loss (CPALoss) to improve the learning and understanding of the encoder for polarization modal information and to optimize the PGA module. With extensive experiments on a total of three RGB-P benchmarks, our ShareCMP achieves state-of-the-art performance in mIoU with fewer parameters on the UPLight (92.45(+0.32)%), ZJU (92.7(+0.1)%), and MCubeS (50.99(+1.51)%) datasets compared to the previous best methods. The code is available at https://github.com/LEFTeyex/ShareCMP.
Authors: Jialong Zuo, Hanyu Zhou, Ying Nie, Feng Zhang, Tianyu Guo, Nong Sang, Yunhe Wang, Changxin Gao
Existing text-based person retrieval datasets often have relatively coarse-grained text annotations. This hinders the model to comprehend the fine-grained semantics of query texts in real scenarios. To address this problem, we contribute a new benchmark named \textbf{UFineBench} for text-based person retrieval with ultra-fine granularity.
Firstly, we construct a new \textbf{dataset} named UFine6926. We collect a large number of person images and manually annotate each image with two detailed textual descriptions, averaging 80.8 words each. The average word count is three to four times that of the previous datasets. In addition of standard in-domain evaluation, we also propose a special \textbf{evaluation paradigm} more representative of real scenarios. It contains a new evaluation set with cross domains, cross textual granularity and cross textual styles, named UFine3C, and a new evaluation metric for accurately measuring retrieval ability, named mean Similarity Distribution (mSD). Moreover, we propose CFAM, a more efficient \textbf{algorithm} especially designed for text-based person retrieval with ultra fine-grained texts. It achieves fine granularity mining by adopting a shared cross-modal granularity decoder and hard negative match mechanism.
With standard in-domain evaluation, CFAM establishes competitive performance across various datasets, especially on our ultra fine-grained UFine6926. Furthermore, by evaluating on UFine3C, we demonstrate that training on our UFine6926 significantly improves generalization to real scenarios compared with other coarse-grained datasets. The dataset and code will be made publicly available at \url{https://github.com/Zplusdragon/UFineBench}.
Authors: Vladimir Arkhipkin, Andrei Filatov, Viacheslav Vasilev, Anastasia Maltseva, Said Azizov, Igor Pavlov, Julia Agafonova, Andrey Kuznetsov, Denis Dimitrov
We present Kandinsky 3.0, a large-scale text-to-image generation model based on latent diffusion, continuing the series of text-to-image Kandinsky models and reflecting our progress to achieve higher quality and realism of image generation. Compared to previous versions of Kandinsky 2.x, Kandinsky 3.0 leverages a two times larger U-Net backbone, a ten times larger text encoder and removes diffusion mapping. We describe the architecture of the model, the data collection procedure, the training technique, and the production system of user interaction. We focus on the key components that, as we have identified as a result of a large number of experiments, had the most significant impact on improving the quality of our model compared to the others. By our side-by-side comparisons, Kandinsky becomes better in text understanding and works better on specific domains. Project page: https://ai-forever.github.io/Kandinsky-3
Authors: Xiaobo Yang, Xiaojin Gong
This work aims to leverage pre-trained foundation models, such as contrastive language-image pre-training (CLIP) and segment anything model (SAM), to address weakly supervised semantic segmentation (WSSS) using image-level labels. To this end, we propose a coarse-to-fine framework based on CLIP and SAM for generating high-quality segmentation seeds. Specifically, we construct an image classification task and a seed segmentation task, which are jointly performed by CLIP with frozen weights and two sets of learnable task-specific prompts. A SAM-based seeding (SAMS) module is designed and applied to each task to produce either coarse or fine seed maps. Moreover, we design a multi-label contrastive loss supervised by image-level labels and a CAM activation loss supervised by the generated coarse seed map. These losses are used to learn the prompts, which are the only parts need to be learned in our framework. Once the prompts are learned, we input each image along with the learned segmentation-specific prompts into CLIP and the SAMS module to produce high-quality segmentation seeds. These seeds serve as pseudo labels to train an off-the-shelf segmentation network like other two-stage WSSS methods. Experiments show that our method achieves the state-of-the-art performance on PASCAL VOC 2012 and competitive results on MS COCO 2014. Code is available at https://github.com/HAL-42/FMA-WSSS.git.
Authors: Yushi Lan, Feitong Tan, Di Qiu, Qiangeng Xu, Kyle Genova, Zeng Huang, Sean Fanello, Rohit Pandey, Thomas Funkhouser, Chen Change Loy, Yinda Zhang
We present a novel framework for generating photorealistic 3D human head and subsequently manipulating and reposing them with remarkable flexibility. The proposed approach leverages an implicit function representation of 3D human heads, employing 3D Gaussians anchored on a parametric face model. To enhance representational capabilities and encode spatial information, we embed a lightweight tri-plane payload within each Gaussian rather than directly storing color and opacity. Additionally, we parameterize the Gaussians in a 2D UV space via a 3DMM, enabling effective utilization of the diffusion model for 3D head avatar generation. Our method facilitates the creation of diverse and realistic 3D human heads with fine-grained editing over facial features and expressions. Extensive experiments demonstrate the effectiveness of our method.
Authors: Boyang Deng
Generative models have increasingly impacted relative tasks, from computer vision to interior design and other fields. Stable diffusion is an outstanding diffusion model that paves the way for producing high-resolution images with thorough details from text prompts or reference images. It will be an interesting topic about gaining improvements for small datasets with image-sparse categories. This study utilized seven common categories and three widespread weed species to evaluate the efficiency of a stable diffusion model. In detail, Stable diffusion was used to generate synthetic images belonging to these classes; three techniques (i.e., Image-to-image translation, Dreambooth, and ControlNet) based on stable diffusion were leveraged for image generation with different focuses. Then, classification and detection tasks were conducted based on these synthetic images, whose performance was compared to the models trained on original images. Promising results have been achieved in some classes. This seminal study may expedite the adaption of stable diffusion models to different fields.
Authors: Kairui Yang, Zihao Guo, Gengjie Lin, Haotian Dong, Die Zuo, Jibin Peng, Zhao Huang, Zhecheng Xu, Fupeng Li, Ziyun Bai, Di Lin
We advocate the idea of the natural-language-driven(NLD) simulation to efficiently produce the object interactions between multiple objects in the virtual road scenes, for teaching and testing the autonomous driving systems that should take quick action to avoid collision with obstacles with unpredictable motions. The NLD simulation allows the brief natural-language description to control the object interactions, significantly reducing the human efforts for creating a large amount of interaction data. To facilitate the research of NLD simulation, we collect the Language-to-Interaction(L2I) benchmark dataset with 120,000 natural-language descriptions of object interactions in 6 common types of road topologies. Each description is associated with the programming code, which the graphic render can use to visually reconstruct the object interactions in the virtual scenes. As a methodology contribution, we design SimCopilot to translate the interaction descriptions to the renderable code. We use the L2I dataset to evaluate SimCopilot's abilities to control the object motions, generate complex interactions, and generalize interactions across road topologies. The L2I dataset and the evaluation results motivate the relevant research of the NLD simulation.
Authors: Thomas Westfechtel, Dexuan Zhang, Tatsuya Harada
Unsupervised domain adaptation (UDA) tries to overcome the tedious work of labeling data by leveraging a labeled source dataset and transferring its knowledge to a similar but different target dataset. On the other hand, current vision-language models exhibit astonishing zero-shot prediction capabilities. In this work, we combine knowledge gained through UDA with the inherent knowledge of vision-language models. In a first step, we generate the zero-shot predictions of the source and target dataset using the vision-language model. Since zero-shot predictions usually exhibit a large entropy, meaning that the class probabilities are rather evenly distributed, we first adjust the distribution to accentuate the winning probabilities. This is done using both source and target data to keep the relative confidence between source and target data. We then employ a conventional DA method, to gain the knowledge from the source dataset, in combination with self-knowledge distillation, to maintain the inherent knowledge of the vision-language model. We further combine our method with a gradual source domain expansion strategy (GSDE) and show that this strategy can also benefit by including zero-shot predictions. We conduct experiments and ablation studies on three benchmarks (OfficeHome, VisDA, and DomainNet) and outperform state-of-the-art methods. We further show in ablation studies the contributions of different parts of our algorithm.
Authors: Yiduo Hao, Sohrab Madani, Junfeng Guan, Mohammed Alloulah, Saurabh Gupta, Haitham Hassanieh
The perception of autonomous vehicles using radars has attracted increased research interest due its ability to operate in fog and bad weather. However, training radar models is hindered by the cost and difficulty of annotating large-scale radar data. To overcome this bottleneck, we propose a self-supervised learning framework to leverage the large amount of unlabeled radar data to pre-train radar-only embeddings for self-driving perception tasks. The proposed method combines radar-to-radar and radar-to-vision contrastive losses to learn a general representation from unlabeled radar heatmaps paired with their corresponding camera images. When used for downstream object detection, we demonstrate that the proposed self-supervision framework can improve the accuracy of state-of-the-art supervised baselines by 5.8% in mAP.
Authors: Yiming Li, Mingyan Zhu, Junfeng Guo, Tao Wei, Shu-Tao Xia, Zhan Qin
Currently, sample-specific backdoor attacks (SSBAs) are the most advanced and malicious methods since they can easily circumvent most of the current backdoor defenses. In this paper, we reveal that SSBAs are not sufficiently stealthy due to their poisoned-label nature, where users can discover anomalies if they check the image-label relationship. In particular, we demonstrate that it is ineffective to directly generalize existing SSBAs to their clean-label variants by poisoning samples solely from the target class. We reveal that it is primarily due to two reasons, including \textbf{(1)} the `antagonistic effects' of ground-truth features and \textbf{(2)} the learning difficulty of sample-specific features. Accordingly, trigger-related features of existing SSBAs cannot be effectively learned under the clean-label setting due to their mild trigger intensity required for ensuring stealthiness. We argue that the intensity constraint of existing SSBAs is mostly because their trigger patterns are `content-irrelevant' and therefore act as `noises' for both humans and DNNs. Motivated by this understanding, we propose to exploit content-relevant features, $a.k.a.$ (human-relied) attributes, as the trigger patterns to design clean-label SSBAs. This new attack paradigm is dubbed backdoor attack with attribute trigger (BAAT). Extensive experiments are conducted on benchmark datasets, which verify the effectiveness of our BAAT and its resistance to existing defenses.
Authors: Mengyang Feng, Jinlin Liu, Kai Yu, Yuan Yao, Zheng Hui, Xiefan Guo, Xianhui Lin, Haolan Xue, Chen Shi, Xiaowen Li, Aojie Li, Xiaoyang Kang, Biwen Lei, Miaomiao Cui, Peiran Ren, Xuansong Xie
In this paper, we present DreaMoving, a diffusion-based controllable video generation framework to produce high-quality customized human videos. Specifically, given target identity and posture sequences, DreaMoving can generate a video of the target identity moving or dancing anywhere driven by the posture sequences. To this end, we propose a Video ControlNet for motion-controlling and a Content Guider for identity preserving. The proposed model is easy to use and can be adapted to most stylized diffusion models to generate diverse results. The project page is available at https://dreamoving.github.io/dreamoving
Authors: Wenbo Hu, Yifan Xu, Yi Li, Weiyue Li, Zeyuan Chen, Zhuowen Tu
Vision Language Models (VLMs), which extend Large Language Models (LLM) by incorporating visual understanding capability, have demonstrated significant advancements in addressing open-ended visual question-answering (VQA) tasks. However, these models cannot accurately interpret images infused with text, a common occurrence in real-world scenarios. Standard procedures for extracting information from images often involve learning a fixed set of query embeddings. These embeddings are designed to encapsulate image contexts and are later used as soft prompt inputs in LLMs. Yet, this process is limited to the token count, potentially curtailing the recognition of scenes with text-rich context. To improve upon them, the present study introduces BLIVA: an augmented version of InstructBLIP with Visual Assistant. BLIVA incorporates the query embeddings from InstructBLIP and also directly projects encoded patch embeddings into the LLM, a technique inspired by LLaVA. This approach assists the model to capture intricate details potentially missed during the query decoding process. Empirical evidence demonstrates that our model, BLIVA, significantly enhances performance in processing text-rich VQA benchmarks (up to 17.76% in OCR-VQA benchmark) and in undertaking general (not particularly text-rich) VQA benchmarks (up to 7.9% in Visual Spatial Reasoning benchmark), comparing to our baseline InstructBLIP. BLIVA demonstrates significant capability in decoding real-world images, irrespective of text presence. To demonstrate the broad industry applications enabled by BLIVA, we evaluate the model using a new dataset comprising YouTube thumbnails paired with question-answer sets across 11 diverse categories. For researchers interested in further exploration, our code and models are freely accessible at https://github.com/mlpc-ucsd/BLIVA.