Authors: L. Julian Lechuga Lopez, Tim G. J. Rudner, Farah E. Shamout
Machine learning-aided clinical decision support has the potential to significantly improve patient care. However, existing efforts in this domain for principled quantification of uncertainty have largely been limited to applications of ad-hoc solutions that do not consistently improve reliability. In this work, we consider stochastic neural networks and design a tailor-made multimodal data-driven (M2D2) prior distribution over network parameters. We use simple and scalable Gaussian mean-field variational inference to train a Bayesian neural network using the M2D2 prior. We train and evaluate the proposed approach using clinical time-series data in MIMIC-IV and corresponding chest X-ray images in MIMIC-CXR for the classification of acute care conditions. Our empirical results show that the proposed method produces a more reliable predictive model compared to deterministic and Bayesian neural network baselines.
Authors: Mert Ege, Mustafa Ceyhan
Online exams are more attractive after the Covid-19 pandemic. Furthermore, during recruitment, online exams are used. However, there are more cheating possibilities for online exams. Assigning a proctor for each exam increases cost. At this point, automatic proctor systems detect possible cheating status. This article proposes an end-to-end system and submodules to get better results for online proctoring. Object detection, face recognition, human voice detection, and segmentation are used in our system. Furthermore, our proposed model works on the PCs of users, meaning a client-based system. So, server cost is eliminated. As far as we know, it is the first time the client-based online proctoring system has been used for recruitment. Online exams are more attractive after the Covid-19 pandemic. Furthermore, during recruitment, online exams are used. However, there are more cheating possibilities for online exams. Assigning a proctor for each exam increases cost. At this point, automatic proctor systems detect possible cheating status. This article proposes an end-to-end system and submodules to get better results for online proctoring. Object detection, face recognition, human voice detection, and segmentation are used in our system. Furthermore, our proposed model works on the PCs of users, meaning a client-based system. So, server cost is eliminated. As far as we know, it is the first time the client-based online proctoring system has been used for recruitment. Furthermore, this cheating system works at https://www.talent-interview.com/tr/.
Authors: Gyanendar Manohar, Ruairi O'Reilly
Glaucoma is an irreversible ocular disease and is the second leading cause of visual disability worldwide. Slow vision loss and the asymptomatic nature of the disease make its diagnosis challenging. Early detection is crucial for preventing irreversible blindness. Ophthalmologists primarily use retinal fundus images as a non-invasive screening method. Convolutional neural networks (CNN) have demonstrated high accuracy in the classification of medical images. Nevertheless, CNN's translation-invariant nature and inability to handle the part-whole relationship between objects make its direct application unsuitable for glaucomatous fundus image classification, as it requires a large number of labelled images for training. This work reviews existing state of the art models and proposes InceptionCaps, a novel capsule network (CapsNet) based deep learning model having pre-trained InceptionV3 as its convolution base, for automatic glaucoma classification. InceptionCaps achieved an accuracy of 0.956, specificity of 0.96, and AUC of 0.9556, which surpasses several state-of-the-art deep learning model performances on the RIM-ONE v2 dataset. The obtained result demonstrates the robustness of the proposed deep learning model.
Authors: Zongqian Wu, Yujing Liu, Mengmeng Zhan, Jialie Shen, Ping Hu, Xiaofeng Zhu
Although current prompt learning methods have successfully been designed to effectively reuse the large pre-trained models without fine-tuning their large number of parameters, they still have limitations to be addressed, i.e., without considering the adverse impact of meaningless patches in every image and without simultaneously considering in-sample generalization and out-of-sample generalization. In this paper, we propose an adaptive multi-modality prompt learning to address the above issues. To do this, we employ previous text prompt learning and propose a new image prompt learning. The image prompt learning achieves in-sample and out-of-sample generalization, by first masking meaningless patches and then padding them with the learnable parameters and the information from texts. Moreover, each of the prompts provides auxiliary information to each other, further strengthening these two kinds of generalization. Experimental results on real datasets demonstrate that our method outperforms SOTA methods, in terms of different downstream tasks.
Authors: Mehmet Can Yavuz, Berrin Yanikoglu
Learning a discriminative semantic space using unlabelled and noisy data remains unaddressed in a multi-label setting. We present a contrastive self-supervised learning method which is robust to data noise, grounded in the domain of variational methods. The method (VCL) utilizes variational contrastive learning with beta-divergence to learn robustly from unlabelled datasets, including uncurated and noisy datasets. We demonstrate the effectiveness of the proposed method through rigorous experiments including linear evaluation and fine-tuning scenarios with multi-label datasets in the face understanding domain. In almost all tested scenarios, VCL surpasses the performance of state-of-the-art self-supervised methods, achieving a noteworthy increase in accuracy.
Authors: Phillip Howard, Avinash Madasu, Tiep Le, Gustavo Lujan Moreno, Anahita Bhiwandiwalla, Vasudev Lal
While vision-language models (VLMs) have achieved remarkable performance improvements recently, there is growing evidence that these models also posses harmful biases with respect to social attributes such as gender and race. Prior studies have primarily focused on probing such bias attributes individually while ignoring biases associated with intersections between social attributes. This could be due to the difficulty of collecting an exhaustive set of image-text pairs for various combinations of social attributes. To address this challenge, we employ text-to-image diffusion models to produce counterfactual examples for probing intserctional social biases at scale. Our approach utilizes Stable Diffusion with cross attention control to produce sets of counterfactual image-text pairs that are highly similar in their depiction of a subject (e.g., a given occupation) while differing only in their depiction of intersectional social attributes (e.g., race & gender). Through our over-generate-then-filter methodology, we produce SocialCounterfactuals, a high-quality dataset containing over 171k image-text pairs for probing intersectional biases related to gender, race, and physical characteristics. We conduct extensive experiments to demonstrate the usefulness of our generated dataset for probing and mitigating intersectional social biases in state-of-the-art VLMs.
Authors: Kyungho Bae, Geo Ahn, Youngrae Kim, Jinwoo Choi
When watching a video, humans can naturally extract human actions from the surrounding scene context, even when action-scene combinations are unusual. However, unlike humans, video action recognition models often learn scene-biased action representations from the spurious correlation in training data, leading to poor performance in out-of-context scenarios. While scene-debiased models achieve improved performance in out-of-context scenarios, they often overlook valuable scene information in the data. Addressing this challenge, we propose Disentangled VIdeo representations of Action and Scene (DEVIAS), which aims to achieve holistic video understanding. Disentangled action and scene representations with our method could provide flexibility to adjust the emphasis on action or scene information depending on downstream task and dataset characteristics. Disentangled action and scene representations could be beneficial for both in-context and out-of-context video understanding. To this end, we employ slot attention to learn disentangled action and scene representations with a single model, along with auxiliary tasks that further guide slot attention. We validate the proposed method on both in-context datasets: UCF-101 and Kinetics-400, and out-of-context datasets: SCUBA and HAT. Our proposed method shows favorable performance across different datasets compared to the baselines, demonstrating its effectiveness in diverse video understanding scenarios.
Authors: Siqi Wang, Chau Pham, Bryan A. Plummer
Noisy labels can impair model performance, making the study of learning with noisy labels an important topic. Two conventional approaches are noise modeling and noise detection. However, these two methods are typically studied independently, and there has been limited work on their collaboration. In this work, we explore the integration of these two approaches, proposing an interconnected structure with three crucial blocks: noise modeling, source knowledge identification, and enhanced noise detection using noise source-knowledge-integration methods. This collaboration structure offers advantages such as discriminating hard negatives and preserving genuinely clean labels that might be suspiciously noisy. Our experiments on four datasets, featuring three types of noise and different combinations of each block, demonstrate the efficacy of these components' collaboration. Our collaborative structure methods achieve up to a 10% increase in top-1 classification accuracy in synthesized noise datasets and 3-5% in real-world noisy datasets. The results also suggest that these components make distinct contributions to overall performance across various noise scenarios. These findings provide valuable insights for designing noisy label learning methods customized for specific noise scenarios in the future. Our code is accessible to the public.
Authors: Dina Bashkirova, Arijit Ray, Rupayan Mallick, Sarah Adel Bargal, Jianming Zhang, Ranjay Krishna, Kate Saenko
Professional artists, photographers, and other visual content creators use object relighting to establish their photo's desired effect. Unfortunately, manual tools that allow relighting have a steep learning curve and are difficult to master. Although generative editing methods now enable some forms of image editing, relighting is still beyond today's capabilities; existing methods struggle to keep other aspects of the image -- colors, shapes, and textures -- consistent after the edit. We propose Lasagna, a method that enables intuitive text-guided relighting control. Lasagna learns a lighting prior by using score distillation sampling to distill the prior of a diffusion model, which has been finetuned on synthetic relighting data. To train Lasagna, we curate a new synthetic dataset ReLiT, which contains 3D object assets re-lit from multiple light source locations. Despite training on synthetic images, quantitative results show that Lasagna relights real-world images while preserving other aspects of the input image, outperforming state-of-the-art text-guided image editing methods. Lasagna enables realistic and controlled results on natural images and digital art pieces and is preferred by humans over other methods in over 91% of cases. Finally, we demonstrate the versatility of our learning objective by extending it to allow colorization, another form of image editing.
Authors: Anton Ratnarajah, Sreyan Ghosh, Sonal Kumar, Purva Chiniya, Dinesh Manocha
Accurate estimation of Room Impulse Response (RIR), which captures an environment's acoustic properties, is important for speech processing and AR/VR applications. We propose AV-RIR, a novel multi-modal multi-task learning approach to accurately estimate the RIR from a given reverberant speech signal and the visual cues of its corresponding environment. AV-RIR builds on a novel neural codec-based architecture that effectively captures environment geometry and materials properties and solves speech dereverberation as an auxiliary task by using multi-task learning. We also propose Geo-Mat features that augment material information into visual cues and CRIP that improves late reverberation components in the estimated RIR via image-to-RIR retrieval by 86%. Empirical results show that AV-RIR quantitatively outperforms previous audio-only and visual-only approaches by achieving 36% - 63% improvement across various acoustic metrics in RIR estimation. Additionally, it also achieves higher preference scores in human evaluation. As an auxiliary benefit, dereverbed speech from AV-RIR shows competitive performance with the state-of-the-art in various spoken language processing tasks and outperforms reverberation time error score in the real-world AVSpeech dataset. Qualitative examples of both synthesized reverberant speech and enhanced speech can be found at https://www.youtube.com/watch?v=tTsKhviukAE.
Authors: Xiaoran Zhang, Daniel H. Pak, Shawn S. Ahn, Xiaoxiao Li, Chenyu You, Lawrence Staib, Albert J. Sinusas, Alex Wong, James S. Duncan
This paper proposes a heteroscedastic uncertainty estimation framework for unsupervised medical image registration. Existing methods rely on objectives (e.g. mean-squared error) that assume a uniform noise level across the image, disregarding the heteroscedastic and input-dependent characteristics of noise distribution in real-world medical images. This further introduces noisy gradients due to undesired penalization on outliers, causing unnatural deformation and performance degradation. To mitigate this, we propose an adaptive weighting scheme with a relative $\gamma$-exponentiated signal-to-noise ratio (SNR) for the displacement estimator after modeling the heteroscedastic noise using a separate variance estimator to prevent the model from being driven away by spurious gradients from error residuals, leading to more accurate displacement estimation. To illustrate the versatility and effectiveness of the proposed method, we tested our framework on two representative registration architectures across three medical image datasets. Our proposed framework consistently outperforms other baselines both quantitatively and qualitatively while also providing accurate and sensible uncertainty measures. Paired t-tests show that our improvements in registration accuracy are statistically significant. The code will be publicly available at \url{https://voldemort108x.github.io/hetero_uncertainty/}.
Authors: Xiaoran Zhang, John C. Stendahl, Lawrence Staib, Albert J. Sinusas, Alex Wong, James S. Duncan
We propose an adaptive training scheme for unsupervised medical image registration. Existing methods rely on image reconstruction as the primary supervision signal. However, nuisance variables (e.g. noise and covisibility) often cause the loss of correspondence between medical images, violating the Lambertian assumption in physical waves (e.g. ultrasound) and consistent imaging acquisition. As the unsupervised learning scheme relies on intensity constancy to establish correspondence between images for reconstruction, this introduces spurious error residuals that are not modeled by the typical training objective. To mitigate this, we propose an adaptive framework that re-weights the error residuals with a correspondence scoring map during training, preventing the parametric displacement estimator from drifting away due to noisy gradients, which leads to performance degradations. To illustrate the versatility and effectiveness of our method, we tested our framework on three representative registration architectures across three medical image datasets along with other baselines. Our proposed adaptive framework consistently outperforms other methods both quantitatively and qualitatively. Paired t-tests show that our improvements are statistically significant. The code will be publicly available at \url{https://voldemort108x.github.io/AdaCS/}.
Authors: Cheng Chen, Jingkuan Song, LianLi Gao, Heng Tao Shen
Catastrophic Forgetting (CF) is a prominent issue in continual learning. Parameter isolation addresses this challenge by masking a sub-network for each task to mitigate interference with old tasks. However, these sub-networks are constructed relying on weight magnitude, which does not necessarily correspond to the importance of weights, resulting in maintaining unimportant weights and constructing redundant sub-networks. To overcome this limitation, inspired by information bottleneck, which removes redundancy between adjacent network layers, we propose \textbf{\underline{I}nformation \underline{B}ottleneck \underline{M}asked sub-network (IBM)} to eliminate redundancy within sub-networks. Specifically, IBM accumulates valuable information into essential weights to construct redundancy-free sub-networks, not only effectively mitigating CF by freezing the sub-networks but also facilitating new tasks training through the transfer of valuable knowledge. Additionally, IBM decomposes hidden representations to automate the construction process and make it flexible. Extensive experiments demonstrate that IBM consistently outperforms state-of-the-art methods. Notably, IBM surpasses the state-of-the-art parameter isolation method with a 70\% reduction in the number of parameters within sub-networks and an 80\% decrease in training time.
Authors: Huadong Li, Minhao Jing, Jiajun Liang, Haoqiang Fan, Renhe Ji
It is widely believed that the dense supervision is better than the sparse supervision in the field of depth completion, but the underlying reasons for this are rarely discussed. In this paper, we find that the challenge of using sparse supervision for training Radar-Camera depth prediction models is the Projection Transformation Collapse (PTC). The PTC implies that sparse supervision leads the model to learn unexpected collapsed projection transformations between Image/Radar/LiDAR spaces. Building on this insight, we propose a novel ``Disruption-Compensation" framework to handle the PTC, thereby relighting the use of sparse supervision in depth completion tasks. The disruption part deliberately discards position correspondences among Image/Radar/LiDAR, while the compensation part leverages 3D spatial and 2D semantic information to compensate for the discarded beneficial position correspondence. Extensive experimental results demonstrate that our framework (sparse supervision) outperforms the state-of-the-art (dense supervision) with 11.6$\%$ improvement in mean absolute error and $1.6 \times$ speedup. The code is available at ...
Authors: Hyeonho Jeong, Geon Yeong Park, Jong Chul Ye
Text-to-video diffusion models have advanced video generation significantly. However, customizing these models to generate videos with tailored motions presents a substantial challenge. In specific, they encounter hurdles in (a) accurately reproducing motion from a target video, and (b) creating diverse visual variations. For example, straightforward extensions of static image customization methods to video often lead to intricate entanglements of appearance and motion data. To tackle this, here we present the Video Motion Customization (VMC) framework, a novel one-shot tuning approach crafted to adapt temporal attention layers within video diffusion models. Our approach introduces a novel motion distillation objective using residual vectors between consecutive frames as a motion reference. The diffusion process then preserves low-frequency motion trajectories while mitigating high-frequency motion-unrelated noise in image space. We validate our method against state-of-the-art video generative models across diverse real-world motions and contexts. Our codes, data and the project demo can be found at https://video-motion-customization.github.io
Authors: Hanlin Chen, Chen Li, Gim Hee Lee
Existing neural implicit surface reconstruction methods have achieved impressive performance in multi-view 3D reconstruction by leveraging explicit geometry priors such as depth maps or point clouds as regularization. However, the reconstruction results still lack fine details because of the over-smoothed depth map or sparse point cloud. In this work, we propose a neural implicit surface reconstruction pipeline with guidance from 3D Gaussian Splatting to recover highly detailed surfaces. The advantage of 3D Gaussian Splatting is that it can generate dense point clouds with detailed structure. Nonetheless, a naive adoption of 3D Gaussian Splatting can fail since the generated points are the centers of 3D Gaussians that do not necessarily lie on the surface. We thus introduce a scale regularizer to pull the centers close to the surface by enforcing the 3D Gaussians to be extremely thin. Moreover, we propose to refine the point cloud from 3D Gaussians Splatting with the normal priors from the surface predicted by neural implicit models instead of using a fixed set of points as guidance. Consequently, the quality of surface reconstruction improves from the guidance of the more accurate 3D Gaussian splatting. By jointly optimizing the 3D Gaussian Splatting and the neural implicit model, our approach benefits from both representations and generates complete surfaces with intricate details. Experiments on Tanks and Temples verify the effectiveness of our proposed method.
Authors: Tianyu Yu, Yuan Yao, Haoye Zhang, Taiwen He, Yifeng Han, Ganqu Cui, Jinyi Hu, Zhiyuan Liu, Hai-Tao Zheng, Maosong Sun, Tat-Seng Chua
Multimodal Large Language Models (MLLMs) have recently demonstrated impressive capabilities in multimodal understanding, reasoning, and interaction. However, existing MLLMs prevalently suffer from serious hallucination problems, generating text that is not factually grounded in associated images. The problem makes existing MLLMs untrustworthy and thus impractical in real-world (especially high-stakes) applications. To address the challenge, we present RLHF-V, which enhances MLLM trustworthiness via behavior alignment from fine-grained correctional human feedback. Specifically, RLHF-V collects human preference in the form of segment-level corrections on hallucinations, and performs dense direct preference optimization over the human feedback. Comprehensive experiments on five benchmarks in both automatic and human evaluation show that, RLHF-V can enable substantially more trustworthy MLLM behaviors with promising data and computation efficiency. Remarkably, using 1.4k annotated data samples, RLHF-V significantly reduces the hallucination rate of the base MLLM by 34.8%, outperforming the concurrent LLaVA-RLHF trained on 10k annotated data. The final model achieves state-of-the-art performance in trustworthiness among open-source MLLMs, and shows better robustness than GPT-4V in preventing hallucinations aroused from over-generalization. We open-source our code, model, and data at https://github.com/RLHF-V/RLHF-V.
Authors: Weiying Xie, Xiaoyi Fan, Xin Zhang, Yunsong Li, Jie Lei, Leyuan Fang
Pruning-quantization joint learning always facilitates the deployment of deep neural networks (DNNs) on resource-constrained edge devices. However, most existing methods do not jointly learn a global criterion for pruning and quantization in an interpretable way. In this paper, we propose a novel physics inspired criterion for pruning-quantization joint learning (PIC-PQ), which is explored from an analogy we first draw between elasticity dynamics (ED) and model compression (MC). Specifically, derived from Hooke's law in ED, we establish a linear relationship between the filters' importance distribution and the filter property (FP) by a learnable deformation scale in the physics inspired criterion (PIC). Furthermore, we extend PIC with a relative shift variable for a global view. To ensure feasibility and flexibility, available maximum bitwidth and penalty factor are introduced in quantization bitwidth assignment. Experiments on benchmarks of image classification demonstrate that PIC-PQ yields a good trade-off between accuracy and bit-operations (BOPs) compression ratio e.g., 54.96X BOPs compression ratio in ResNet56 on CIFAR10 with 0.10% accuracy drop and 53.24X in ResNet18 on ImageNet with 0.61% accuracy drop). The code will be available at https://github.com/fanxxxxyi/PIC-PQ.
Authors: Litu Rout, Yujia Chen, Abhishek Kumar, Constantine Caramanis, Sanjay Shakkottai, Wen-Sheng Chu
Sampling from the posterior distribution poses a major computational challenge in solving inverse problems using latent diffusion models. Common methods rely on Tweedie's first-order moments, which are known to induce a quality-limiting bias. Existing second-order approximations are impractical due to prohibitive computational costs, making standard reverse diffusion processes intractable for posterior sampling. This paper introduces Second-order Tweedie sampler from Surrogate Loss (STSL), a novel sampler that offers efficiency comparable to first-order Tweedie with a tractable reverse process using second-order approximation. Our theoretical results reveal that the second-order approximation is lower bounded by our surrogate loss that only requires $O(1)$ compute using the trace of the Hessian, and by the lower bound we derive a new drift term to make the reverse process tractable. Our method surpasses SoTA solvers PSLD and P2L, achieving 4X and 8X reduction in neural function evaluations, respectively, while notably enhancing sampling quality on FFHQ, ImageNet, and COCO benchmarks. In addition, we show STSL extends to text-guided image editing and addresses residual distortions present from corrupted images in leading text-guided image editing methods. To our best knowledge, this is the first work to offer an efficient second-order approximation in solving inverse problems using latent diffusion and editing real-world images with corruptions.
Authors: Xi Yang, Chenhang He, Jianqi Ma, Lei Zhang
Real-world low-resolution (LR) videos have diverse and complex degradations, imposing great challenges on video super-resolution (VSR) algorithms to reproduce their high-resolution (HR) counterparts with high quality. Recently, the diffusion models have shown compelling performance in generating realistic details for image restoration tasks. However, the diffusion process has randomness, making it hard to control the contents of restored images. This issue becomes more serious when applying diffusion models to VSR tasks because temporal consistency is crucial to the perceptual quality of videos. In this paper, we propose an effective real-world VSR algorithm by leveraging the strength of pre-trained latent diffusion models. To ensure the content consistency among adjacent frames, we exploit the temporal dynamics in LR videos to guide the diffusion process by optimizing the latent sampling path with a motion-guided loss, ensuring that the generated HR video maintains a coherent and continuous visual flow. To further mitigate the discontinuity of generated details, we insert temporal module to the decoder and fine-tune it with an innovative sequence-oriented loss. The proposed motion-guided latent diffusion (MGLD) based VSR algorithm achieves significantly better perceptual quality than state-of-the-arts on real-world VSR benchmark datasets, validating the effectiveness of the proposed model design and training strategies.
Authors: Shuchao Duan, Amirhossein Dadashzadeh, Alan Whone, Majid Mirmehdi
Facial expression recognition (FER) methods have made great inroads in categorising moods and feelings in humans. Beyond FER, pain estimation methods assess levels of intensity in pain expressions, however assessing the quality of all facial expressions is of critical value in health-related applications. In this work, we address the quality of five different facial expressions in patients affected by Parkinson's disease. We propose a novel landmark-guided approach, QAFE-Net, that combines temporal landmark heatmaps with RGB data to capture small facial muscle movements that are encoded and mapped to severity scores. The proposed approach is evaluated on a new Parkinson's Disease Facial Expression dataset (PFED5), as well as on the pain estimation benchmark, the UNBC-McMaster Shoulder Pain Expression Archive Database. Our comparative experiments demonstrate that the proposed method outperforms SOTA action quality assessment works on PFED5 and achieves lower mean absolute error than the SOTA pain estimation methods on UNBC-McMaster. Our code and the new PFED5 dataset are available at https://github.com/shuchaoduan/QAFE-Net.
Authors: Xinyin Ma, Gongfan Fang, Xinchao Wang
Diffusion models have recently gained unprecedented attention in the field of image synthesis due to their remarkable generative capabilities. Notwithstanding their prowess, these models often incur substantial computational costs, primarily attributed to the sequential denoising process and cumbersome model size. Traditional methods for compressing diffusion models typically involve extensive retraining, presenting cost and feasibility challenges. In this paper, we introduce DeepCache, a novel training-free paradigm that accelerates diffusion models from the perspective of model architecture. DeepCache capitalizes on the inherent temporal redundancy observed in the sequential denoising steps of diffusion models, which caches and retrieves features across adjacent denoising stages, thereby curtailing redundant computations. Utilizing the property of the U-Net, we reuse the high-level features while updating the low-level features in a very cheap way. This innovative strategy, in turn, enables a speedup factor of 2.3$\times$ for Stable Diffusion v1.5 with only a 0.05 decline in CLIP Score, and 4.1$\times$ for LDM-4-G with a slight decrease of 0.22 in FID on ImageNet. Our experiments also demonstrate DeepCache's superiority over existing pruning and distillation methods that necessitate retraining and its compatibility with current sampling techniques. Furthermore, we find that under the same throughput, DeepCache effectively achieves comparable or even marginally improved results with DDIM or PLMS. The code is available at https://github.com/horseee/DeepCache
Authors: Jiazhong Cen, Jiemin Fang, Chen Yang, Lingxi Xie, Xiaopeng Zhang, Wei Shen, Qi Tian
Interactive 3D segmentation in radiance fields is an appealing task since its importance in 3D scene understanding and manipulation. However, existing methods face challenges in either achieving fine-grained, multi-granularity segmentation or contending with substantial computational overhead, inhibiting real-time interaction. In this paper, we introduce Segment Any 3D GAussians (SAGA), a novel 3D interactive segmentation approach that seamlessly blends a 2D segmentation foundation model with 3D Gaussian Splatting (3DGS), a recent breakthrough of radiance fields. SAGA efficiently embeds multi-granularity 2D segmentation results generated by the segmentation foundation model into 3D Gaussian point features through well-designed contrastive training. Evaluation on existing benchmarks demonstrates that SAGA can achieve competitive performance with state-of-the-art methods. Moreover, SAGA achieves multi-granularity segmentation and accommodates various prompts, including points, scribbles, and 2D masks. Notably, SAGA can finish the 3D segmentation within milliseconds, achieving nearly 1000x acceleration compared to previous SOTA. The project page is at https://jumpat.github.io/SAGA.
Authors: Yunyang Xiong, Bala Varadarajan, Lemeng Wu, Xiaoyu Xiang, Fanyi Xiao, Chenchen Zhu, Xiaoliang Dai, Dilin Wang, Fei Sun, Forrest Iandola, Raghuraman Krishnamoorthi, Vikas Chandra
Segment Anything Model (SAM) has emerged as a powerful tool for numerous vision applications. A key component that drives the impressive performance for zero-shot transfer and high versatility is a super large Transformer model trained on the extensive high-quality SA-1B dataset. While beneficial, the huge computation cost of SAM model has limited its applications to wider real-world applications. To address this limitation, we propose EfficientSAMs, light-weight SAM models that exhibits decent performance with largely reduced complexity. Our idea is based on leveraging masked image pretraining, SAMI, which learns to reconstruct features from SAM image encoder for effective visual representation learning. Further, we take SAMI-pretrained light-weight image encoders and mask decoder to build EfficientSAMs, and finetune the models on SA-1B for segment anything task. We perform evaluations on multiple vision tasks including image classification, object detection, instance segmentation, and semantic object detection, and find that our proposed pretraining method, SAMI, consistently outperforms other masked image pretraining methods. On segment anything task such as zero-shot instance segmentation, our EfficientSAMs with SAMI-pretrained lightweight image encoders perform favorably with a significant gain (e.g., ~4 AP on COCO/LVIS) over other fast SAM models.
Authors: Xiaoke Huang, Jianfeng Wang, Yansong Tang, Zheng Zhang, Han Hu, Jiwen Lu, Lijuan Wang, Zicheng Liu
We propose a method to efficiently equip the Segment Anything Model (SAM) with the ability to generate regional captions. SAM presents strong generalizability to segment anything while is short for semantic understanding. By introducing a lightweight query-based feature mixer, we align the region-specific features with the embedding space of language models for later caption generation. As the number of trainable parameters is small (typically in the order of tens of millions), it costs less computation, less memory usage, and less communication bandwidth, resulting in both fast and scalable training. To address the scarcity problem of regional caption data, we propose to first pre-train our model on objection detection and segmentation tasks. We call this step weak supervision pretraining since the pre-training data only contains category names instead of full-sentence descriptions. The weak supervision pretraining allows us to leverage many publicly available object detection and segmentation datasets. We conduct extensive experiments to demonstrate the superiority of our method and validate each design choice. This work serves as a stepping stone towards scaling up regional captioning data and sheds light on exploring efficient ways to augment SAM with regional semantics. The project page, along with the associated code, can be accessed via the following https://xk-huang.github.io/segment-caption-anything/.
Authors: Balamurugan Thambiraja, Sadegh Aliakbarian, Darren Cosker, Justus Thies
We present 3DiFACE, a novel method for personalized speech-driven 3D facial animation and editing. While existing methods deterministically predict facial animations from speech, they overlook the inherent one-to-many relationship between speech and facial expressions, i.e., there are multiple reasonable facial expression animations matching an audio input. It is especially important in content creation to be able to modify generated motion or to specify keyframes. To enable stochasticity as well as motion editing, we propose a lightweight audio-conditioned diffusion model for 3D facial motion. This diffusion model can be trained on a small 3D motion dataset, maintaining expressive lip motion output. In addition, it can be finetuned for specific subjects, requiring only a short video of the person. Through quantitative and qualitative evaluations, we show that our method outperforms existing state-of-the-art techniques and yields speech-driven animations with greater fidelity and diversity.
Authors: Walid Bousselham, Felix Petersen, Vittorio Ferrari, Hilde Kuehne
Vision-language foundation models have shown remarkable performance in various zero-shot settings such as image retrieval, classification, or captioning. But so far, those models seem to fall behind when it comes to zero-shot localization of referential expressions and objects in images. As a result, they need to be fine-tuned for this task. In this paper, we show that pretrained vision-language (VL) models allow for zero-shot open-vocabulary object localization without any fine-tuning. To leverage those capabilities, we propose a Grounding Everything Module (GEM) that generalizes the idea of value-value attention introduced by CLIPSurgery to a self-self attention path. We show that the concept of self-self attention corresponds to clustering, thus enforcing groups of tokens arising from the same object to be similar while preserving the alignment with the language space. To further guide the group formation, we propose a set of regularizations that allows the model to finally generalize across datasets and backbones. We evaluate the proposed GEM framework on various benchmark tasks and datasets for semantic segmentation. It shows that GEM not only outperforms other training-free open-vocabulary localization methods, but also achieves state-of-the-art results on the recently proposed OpenImagesV7 large-scale segmentation benchmark.
Authors: Botos Csaba, Wenxuan Zhang, Matthias Müller, Ser-Nam Lim, Mohamed Elhoseiny, Philip Torr, Adel Bibi
Online continual learning, the process of training models on streaming data, has gained increasing attention in recent years. However, a critical aspect often overlooked is the label delay, where new data may not be labeled due to slow and costly annotation processes. We introduce a new continual learning framework with explicit modeling of the label delay between data and label streams over time steps. In each step, the framework reveals both unlabeled data from the current time step $t$ and labels delayed with $d$ steps, from the time step $t-d$. In our extensive experiments amounting to 1060 GPU days, we show that merely augmenting the computational resources is insufficient to tackle this challenge. Our findings underline a notable performance decline when solely relying on labeled data when the label delay becomes significant. More surprisingly, when using state-of-the-art SSL and TTA techniques to utilize the newer, unlabeled data, they fail to surpass the performance of a na\"ive method that simply trains on the delayed supervised stream. To this end, we introduce a simple, efficient baseline that rehearses from the labeled memory samples that are most similar to the new unlabeled samples. This method bridges the accuracy gap caused by label delay without significantly increasing computational complexity. We show experimentally that our method is the least affected by the label delay factor and in some cases successfully recovers the accuracy of the non-delayed counterpart. We conduct various ablations and sensitivity experiments, demonstrating the effectiveness of our approach.
Authors: Rohan Choudhury, Koichiro Niinuma, Kris M. Kitani, László A. Jeni
We propose to answer zero-shot questions about videos by generating short procedural programs that derive a final answer from solving a sequence of visual subtasks. We present Procedural Video Querying (ProViQ), which uses a large language model to generate such programs from an input question and an API of visual modules in the prompt, then executes them to obtain the output. Recent similar procedural approaches have proven successful for image question answering, but videos remain challenging: we provide ProViQ with modules intended for video understanding, allowing it to generalize to a wide variety of videos. This code generation framework additionally enables ProViQ to perform other video tasks in addition to question answering, such as multi-object tracking or basic video editing. ProViQ achieves state-of-the-art results on a diverse range of benchmarks, with improvements of up to 25% on short, long, open-ended, and multimodal video question-answering datasets. Our project page is at https://rccchoudhury.github.io/proviq2023.
Authors: Rishi Upadhyay, Howard Zhang, Yunhao Ba, Ethan Yang, Blake Gella, Sicheng Jiang, Alex Wong, Achuta Kadambi
While perspective is a well-studied topic in art, it is generally taken for granted in images. However, for the recent wave of high-quality image synthesis methods such as latent diffusion models, perspective accuracy is not an explicit requirement. Since these methods are capable of outputting a wide gamut of possible images, it is difficult for these synthesized images to adhere to the principles of linear perspective. We introduce a novel geometric constraint in the training process of generative models to enforce perspective accuracy. We show that outputs of models trained with this constraint both appear more realistic and improve performance of downstream models trained on generated images. Subjective human trials show that images generated with latent diffusion models trained with our constraint are preferred over images from the Stable Diffusion V2 model 70% of the time. SOTA monocular depth estimation models such as DPT and PixelFormer, fine-tuned on our images, outperform the original models trained on real images by up to 7.03% in RMSE and 19.3% in SqRel on the KITTI test set for zero-shot transfer.
Authors: Andrea Caraffa, Davide Boscaini, Amir Hamza, Fabio Poiesi
Object 6D pose estimation methods can achieve high accuracy when trained and tested on the same objects. However, estimating the pose of objects that are absent at training time is still a challenge. In this work, we advance the state-of-the-art in zero-shot object 6D pose estimation by proposing the first method that fuses the contribution of pre-trained geometric and vision foundation models. Unlike state-of-the-art approaches that train their pipeline on data specifically crafted for the 6D pose estimation task, our method does not require task-specific finetuning. Instead, our method, which we name PoMZ, combines geometric descriptors learned from point cloud data with visual features learned from large-scale web images to produce distinctive 3D point-level descriptors. By applying an off-the-shelf registration algorithm, like RANSAC, PoMZ outperforms all state-of-the-art zero-shot object 6D pose estimation approaches. We extensively evaluate PoMZ across the seven core datasets of the BOP Benchmark, encompassing over a hundred objects and 20 thousand images captured in diverse scenarios. PoMZ ranks first in the BOP Benchmark under the category Task 4: 6D localization of unseen objects. We will release the source code publicly.
Authors: Kaifeng Chen, Daniel Salz, Huiwen Chang, Kihyuk Sohn, Dilip Krishnan, Mojtaba Seyedhosseini
Training visual embeddings with labeled data supervision has been the de facto setup for representation learning in computer vision. Inspired by recent success of adopting masked image modeling (MIM) in self-supervised representation learning, we propose a simple yet effective setup that can easily integrate MIM into existing supervised training paradigms. In our design, in addition to the original classification task applied to a vision transformer image encoder, we add a shallow transformer-based decoder on top of the encoder and introduce an MIM task which tries to reconstruct image tokens based on masked image inputs. We show with minimal change in architecture and no overhead in inference that this setup is able to improve the quality of the learned representations for downstream tasks such as classification, image retrieval, and semantic segmentation. We conduct a comprehensive study and evaluation of our setup on public benchmarks. On ImageNet-1k, our ViT-B/14 model achieves 81.72% validation accuracy, 2.01% higher than the baseline model. On K-Nearest-Neighbor image retrieval evaluation with ImageNet-1k, the same model outperforms the baseline by 1.32%. We also show that this setup can be easily scaled to larger models and datasets. Code and checkpoints will be released.
Authors: Jialin Wu, Xia Hu, Yaqing Wang, Bo Pang, Radu Soricut
Large multi-modal models (LMMs) exhibit remarkable performance across numerous tasks. However, generalist LMMs often suffer from performance degradation when tuned over a large collection of tasks. Recent research suggests that Mixture of Experts (MoE) architectures are useful for instruction tuning, but for LMMs of parameter size around O(50-100B), the prohibitive cost of replicating and storing the expert models severely limits the number of experts we can use. We propose Omni-SMoLA, an architecture that uses the Soft MoE approach to (softly) mix many multimodal low rank experts, and avoids introducing a significant number of new parameters compared to conventional MoE models. The core intuition here is that the large model provides a foundational backbone, while different lightweight experts residually learn specialized knowledge, either per-modality or multimodally. Extensive experiments demonstrate that the SMoLA approach helps improve the generalist performance across a broad range of generative vision-and-language tasks, achieving new SoTA generalist performance that often matches or outperforms single specialized LMM baselines, as well as new SoTA specialist performance.
Authors: Julian Knodt, Xifeng Gao
Given a 3D mesh with a UV parameterization, we introduce a novel approach to generating textures from text prompts. While prior work uses optimization from Text-to-Image Diffusion models to generate textures and geometry, this is slow and requires significant compute resources. Alternatively, there are projection based approaches that use the same Text-to-Image models that paint images onto a mesh, but lack consistency at different viewing angles, we propose a method that uses a single Depth-to-Image diffusion network, and generates a single consistent texture when rendered on the 3D surface by first unifying multiple 2D image's diffusion paths, and hoisting that to 3D with MultiDiffusion~\cite{multidiffusion}. We demonstrate our approach on a dataset containing 30 meshes, taking approximately 5 minutes per mesh. To evaluate the quality of our approach, we use CLIP-score~\cite{clipscore} and Frechet Inception Distance (FID)~\cite{frechet} to evaluate the quality of the rendering, and show our improvement over prior work.
Authors: An Ngo, MinhPhuong Cao, Rajesh Kumar
While previous studies have explored attacks via random, simple, and skilled forgeries, generative attacks have received limited attention in the data-driven signature verification (DASV) process. Thus, this paper explores the impact of generative attacks on DASV and proposes practical and interpretable countermeasures. We investigate the power of two prominent Deep Generative Models (DGMs), Variational Auto-encoders (VAE) and Conditional Generative Adversarial Networks (CGAN), on their ability to generate signatures that would successfully deceive DASV. Additionally, we evaluate the quality of generated images using the Structural Similarity Index measure (SSIM) and use the same to explain the attack's success. Finally, we propose countermeasures that effectively reduce the impact of deep generative attacks on DASV.
We first generated six synthetic datasets from three benchmark offline-signature datasets viz. CEDAR, BHSig260- Bengali, and BHSig260-Hindi using VAE and CGAN. Then, we built baseline DASVs using Xception, ResNet152V2, and DenseNet201. These DASVs achieved average (over the three datasets) False Accept Rates (FARs) of 2.55%, 3.17%, and 1.06%, respectively. Then, we attacked these baselines using the synthetic datasets. The VAE-generated signatures increased average FARs to 10.4%, 10.1%, and 7.5%, while CGAN-generated signatures to 32.5%, 30%, and 26.1%. The variation in the effectiveness of attack for VAE and CGAN was investigated further and explained by a strong (rho = -0.86) negative correlation between FARs and SSIMs. We created another set of synthetic datasets and used the same to retrain the DASVs. The retained baseline showed significant robustness to random, skilled, and generative attacks as the FARs shrank to less than 1% on average. The findings underscore the importance of studying generative attacks and potential countermeasures for DASV.
Authors: Xiaoyu Wang, Yuchi Ma, Qunying Huang, Zhengwei Yang, Zhou Zhang
Remote sensing technology has become a promising tool in yield prediction. Most prior work employs satellite imagery for county-level corn yield prediction by spatially aggregating all pixels within a county into a single value, potentially overlooking the detailed information and valuable insights offered by more granular data. To this end, this research examines each county at the pixel level and applies multiple instance learning to leverage detailed information within a county. In addition, our method addresses the "mixed pixel" issue caused by the inconsistent resolution between feature datasets and crop mask, which may introduce noise into the model and therefore hinder accurate yield prediction. Specifically, the attention mechanism is employed to automatically assign weights to different pixels, which can mitigate the influence of mixed pixels. The experimental results show that the developed model outperforms four other machine learning models over the past five years in the U.S. corn belt and demonstrates its best performance in 2022, achieving a coefficient of determination (R2) value of 0.84 and a root mean square error (RMSE) of 0.83. This paper demonstrates the advantages of our approach from both spatial and temporal perspectives. Furthermore, through an in-depth study of the relationship between mixed pixels and attention, it is verified that our approach can capture critical feature information while filtering out noise from mixed pixels.
Authors: Jaewoo Jung, Jisang Han, Jiwon Kang, Seongchan Kim, Min-Seop Kwak, Seungryong Kim
Recently, neural radiance field (NeRF) has shown remarkable performance in novel view synthesis and 3D reconstruction. However, it still requires abundant high-quality images, limiting its applicability in real-world scenarios. To overcome this limitation, recent works have focused on training NeRF only with sparse viewpoints by giving additional regularizations, often called few-shot NeRF. We observe that due to the under-constrained nature of the task, solely using additional regularization is not enough to prevent the model from overfitting to sparse viewpoints. In this paper, we propose a novel framework, dubbed Self-Evolving Neural Radiance Fields (SE-NeRF), that applies a self-training framework to NeRF to address these problems. We formulate few-shot NeRF into a teacher-student framework to guide the network to learn a more robust representation of the scene by training the student with additional pseudo labels generated from the teacher. By distilling ray-level pseudo labels using distinct distillation schemes for reliable and unreliable rays obtained with our novel reliability estimation method, we enable NeRF to learn a more accurate and robust geometry of the 3D scene. We show and evaluate that applying our self-training framework to existing models improves the quality of the rendered images and achieves state-of-the-art performance in multiple settings.
Authors: Shentong Mo, Pedro Morgado
Humans possess a remarkable ability to integrate auditory and visual information, enabling a deeper understanding of the surrounding environment. This early fusion of audio and visual cues, demonstrated through cognitive psychology and neuroscience research, offers promising potential for developing multimodal perception models. However, training early fusion architectures poses significant challenges, as the increased model expressivity requires robust learning frameworks to harness their enhanced capabilities. In this paper, we address this challenge by leveraging the masked reconstruction framework, previously successful in unimodal settings, to train audio-visual encoders with early fusion. Additionally, we propose an attention-based fusion module that captures interactions between local audio and visual representations, enhancing the model's ability to capture fine-grained interactions. While effective, this procedure can become computationally intractable, as the number of local representations increases. Thus, to address the computational complexity, we propose an alternative procedure that factorizes the local representations before representing audio-visual interactions. Extensive evaluations on a variety of datasets demonstrate the superiority of our approach in audio-event classification, visual sound localization, sound separation, and audio-visual segmentation. These contributions enable the efficient training of deeply integrated audio-visual models and significantly advance the usefulness of early fusion architectures.
Authors: Minchul Kim, Shangqian Gao, Yen-Chang Hsu, Yilin Shen, Hongxia Jin
Vision Transformers (ViTs) have emerged as powerful backbones in computer vision, outperforming many traditional CNNs. However, their computational overhead, largely attributed to the self-attention mechanism, makes deployment on resource-constrained edge devices challenging. Multiple solutions rely on token pruning or token merging. In this paper, we introduce "Token Fusion" (ToFu), a method that amalgamates the benefits of both token pruning and token merging. Token pruning proves advantageous when the model exhibits sensitivity to input interpolations, while token merging is effective when the model manifests close to linear responses to inputs. We combine this to propose a new scheme called Token Fusion. Moreover, we tackle the limitations of average merging, which doesn't preserve the intrinsic feature norm, resulting in distributional shifts. To mitigate this, we introduce MLERP merging, a variant of the SLERP technique, tailored to merge multiple tokens while maintaining the norm distribution. ToFu is versatile, applicable to ViTs with or without additional training. Our empirical evaluations indicate that ToFu establishes new benchmarks in both classification and image generation tasks concerning computational efficiency and model accuracy.
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.
Authors: Xu Yang, Hanwang Zhang, Jianfei Cai
Dataset bias in vision-language tasks is becoming one of the main problems which hinders the progress of our community. Existing solutions lack a principled analysis about why modern image captioners easily collapse into dataset bias. In this paper, we present a novel perspective: Deconfounded Image Captioning (DIC), to find out the answer of this question, then retrospect modern neural image captioners, and finally propose a DIC framework: DICv1.0 to alleviate the negative effects brought by dataset bias. DIC is based on causal inference, whose two principles: the backdoor and front-door adjustments, help us review previous studies and design new effective models. In particular, we showcase that DICv1.0 can strengthen two prevailing captioning models and can achieve a single-model 131.1 CIDEr-D and 128.4 c40 CIDEr-D on Karpathy split and online split of the challenging MS COCO dataset, respectively. Interestingly, DICv1.0 is a natural derivation from our causal retrospect, which opens promising directions for image captioning.
Authors: Edgar Avalos, Kazuto Akagi, Yasumasa Nishiura
Although the process variables of epoxy resins alter their mechanical properties, the visual identification of the characteristic features of X-ray images of samples of these materials is challenging. To facilitate the identification, we approximate the magnitude of the gradient of the intensity field of the X-ray images of different kinds of epoxy resins and then we use deep learning to discover the most representative features of the transformed images. In this solution of the inverse problem to finding characteristic features to discriminate samples of heterogeneous materials, we use the eigenvectors obtained from the singular value decomposition of all the channels of the feature maps of the early layers in a convolutional neural network. While the strongest activated channel gives a visual representation of the characteristic features, often these are not robust enough in some practical settings. On the other hand, the left singular vectors of the matrix decomposition of the feature maps, barely change when variables such as the capacity of the network or network architecture change. High classification accuracy and robustness of characteristic features are presented in this work.
Authors: Yaxing Wang, Abel Gonzalez-Garcia, Chenshen Wu, Luis Herranz, Fahad Shahbaz Khan, Shangling Jui, Joost van de Weijer
GANs largely increases the potential impact of generative models. Therefore, we propose a novel knowledge transfer method for generative models based on mining the knowledge that is most beneficial to a specific target domain, either from a single or multiple pretrained GANs. This is done using a miner network that identifies which part of the generative distribution of each pretrained GAN outputs samples closest to the target domain. Mining effectively steers GAN sampling towards suitable regions of the latent space, which facilitates the posterior finetuning and avoids pathologies of other methods, such as mode collapse and lack of flexibility. Furthermore, to prevent overfitting on small target domains, we introduce sparse subnetwork selection, that restricts the set of trainable neurons to those that are relevant for the target dataset. We perform comprehensive experiments on several challenging datasets using various GAN architectures (BigGAN, Progressive GAN, and StyleGAN) and show that the proposed method, called MineGAN, effectively transfers knowledge to domains with few target images, outperforming existing methods. In addition, MineGAN can successfully transfer knowledge from multiple pretrained GANs.
Authors: João P C Bertoldo, Etienne Decencière, David Ryckelynck, Henry Proudhon
X-ray Computed Tomography (XCT) techniques have evolved to a point that high-resolution data can be acquired so fast that classic segmentation methods are prohibitively cumbersome, demanding automated data pipelines capable of dealing with non-trivial 3D images. Deep learning has demonstrated success in many image processing tasks, including material science applications, showing a promising alternative for a humanfree segmentation pipeline. In this paper a modular interpretation of UNet (Modular U-Net) is proposed and trained to segment 3D tomography images of a three-phased glass fiber-reinforced Polyamide 66. We compare 2D and 3D versions of our model, finding that the former is slightly better than the latter. We observe that human-comparable results can be achievied even with only 10 annotated layers and using a shallow U-Net yields better results than a deeper one. As a consequence, Neural Network (NN) show indeed a promising venue to automate XCT data processing pipelines needing no human, adhoc intervention.
Authors: Xingqun Qi, Muyi Sun, Zijian Wang, Jiaming Liu, Qi Li, Fang Zhao, Shanghang Zhang, Caifeng Shan
Biphasic face photo-sketch synthesis has significant practical value in wide-ranging fields such as digital entertainment and law enforcement. Previous approaches directly generate the photo-sketch in a global view, they always suffer from the low quality of sketches and complex photo variations, leading to unnatural and low-fidelity results. In this paper, we propose a novel Semantic-Driven Generative Adversarial Network to address the above issues, cooperating with Graph Representation Learning. Considering that human faces have distinct spatial structures, we first inject class-wise semantic layouts into the generator to provide style-based spatial information for synthesized face photos and sketches. Additionally, to enhance the authenticity of details in generated faces, we construct two types of representational graphs via semantic parsing maps upon input faces, dubbed the IntrA-class Semantic Graph (IASG) and the InteR-class Structure Graph (IRSG). Specifically, the IASG effectively models the intra-class semantic correlations of each facial semantic component, thus producing realistic facial details. To preserve the generated faces being more structure-coordinated, the IRSG models inter-class structural relations among every facial component by graph representation learning. To further enhance the perceptual quality of synthesized images, we present a biphasic interactive cycle training strategy by fully taking advantage of the multi-level feature consistency between the photo and sketch. Extensive experiments demonstrate that our method outperforms the state-of-the-art competitors on the CUFS and CUFSF datasets.
Authors: Rintaro Yanagi, Atsushi Hashimoto, Shusaku Sone, Naoya Chiba, Jiaxin Ma, Yoshitaka Ushiku
We present a novel differential matching algorithm for 3D point cloud registration. Instead of only optimizing the feature extractor for a matching algorithm, we propose a learning-based matching module optimized to the jointly-trained feature extractor. We focused on edge-wise feature-forwarding architectures, which are memory-consuming but can avoid the over-smoothing effect that GNNs suffer. We improve its memory efficiency to scale it for point cloud registration while investigating the best way of connecting it to the feature extractor. Experimental results show our matching module's significant impact on performance improvement in rigid/non-rigid and whole/partial point cloud registration datasets with multiple contemporary feature extractors. For example, our module boosted the current SOTA method, RoITr, by +5.4%, and +7.2% in the NFMR metric and +6.1% and +8.5% in the IR metric on the 4DMatch and 4DLoMatch datasets, respectively.
Authors: Iraj Abedi, Mahsa Vali, Bentolhoda Otroshi Shahreza, Hamidreza Bolhasani
During the COVID-19 pandemic, computed tomography (CT) is a good way to diagnose COVID-19 patients. HRCT (High-Resolution Computed Tomography) is a form of computed tomography that uses advanced methods to improve image resolution. Publicly accessible COVID-19 CT image datasets are very difficult to come by due to privacy concerns, which impedes the study and development of AI-powered COVID-19 diagnostic algorithms based on CT images. To address this problem, we have introduced HRCTv1-COVID-19, a new COVID-19 high resolution chest CT Scan image dataset that includes not only COVID-19 cases of Ground Glass Opacity (GGO), Crazy Paving, and Air Space Consolidation, but also CT images of cases with negative COVID-19. The HRCTv1-COVID-19 dataset, which includes slice-level, and patient-level labels, has the potential to aid COVID-19 research, especially for diagnosis and differentiation using artificial intelligence algorithms, machine learning and deep learning methods. This dataset is accessible through web at: this http URL and includes 181,106 chest HRCT images from 395 patients with four labels: GGO, Crazy Paving, Air Space Consolidation and Negative.
Keywords- Dataset, COVID-19, CT-Scan, Computed Tomography, Medical Imaging, Chest Image.
Authors: Joao P. C. Bertoldo, Etienne Decencière
Fully Convolutional Data Description (FCDD), an explainable version of the Hypersphere Classifier (HSC), directly addresses image anomaly detection (AD) and pixel-wise AD without any post-hoc explainer methods. The authors claim that FCDD achieves results comparable with the state-of-the-art in sample-wise AD on Fashion-MNIST and CIFAR-10 and exceeds the state-of-the-art on the pixel-wise task on MVTec-AD. We reproduced the main results of the paper using the author's code with minor changes and provide runtime requirements to achieve if (CPU memory, GPU memory, and training time). We propose another analysis methodology using a critical difference diagram, and further investigate the test performance of the model during the training phase.
Authors: Mingjie Wang, Jianxiong Guo, Sirui Li, Dingwen Xiao, Zhiqing Tang
Deep neural networks have significantly advanced person re-identification (ReID) applications in the realm of the industrial internet, yet they remain vulnerable. Thus, it is crucial to study the robustness of ReID systems, as there are risks of adversaries using these vulnerabilities to compromise industrial surveillance systems. Current adversarial methods focus on generating attack samples using misclassification feedback from victim models (VMs), neglecting VM's cognitive processes. We seek to address this by producing authentic ReID attack instances through VM cognition decryption. This approach boasts advantages like better transferability to open-set ReID tests, easier VM misdirection, and enhanced creation of realistic and undetectable assault images. However, the task of deciphering the cognitive mechanism in VM is widely considered to be a formidable challenge. In this paper, we propose a novel inconspicuous and controllable ReID attack baseline, LCYE (Look Closer to Your Enemy), to generate adversarial query images. Specifically, LCYE first distills VM's knowledge via teacher-student memory mimicking the proxy task. This knowledge prior serves as an unambiguous cryptographic token, encapsulating elements deemed indispensable and plausible by the VM, with the intent of facilitating precise adversarial misdirection. Further, benefiting from the multiple opposing task framework of LCYE, we investigate the interpretability and generalization of ReID models from the view of the adversarial attack, including cross-domain adaption, cross-model consensus, and online learning process. Extensive experiments on four ReID benchmarks show that our method outperforms other state-of-the-art attackers with a large margin in white-box, black-box, and target attacks. The source code can be found at https://github.com/MingjieWang0606/LCYE-attack_reid.
Authors: Lucas Böttcher, Gregory Wheeler
Analyzing geometric properties of high-dimensional loss functions, such as local curvature and the existence of other optima around a certain point in loss space, can help provide a better understanding of the interplay between neural network structure, implementation attributes, and learning performance. In this work, we combine concepts from high-dimensional probability and differential geometry to study how curvature properties in lower-dimensional loss representations depend on those in the original loss space. We show that saddle points in the original space are rarely correctly identified as such in expected lower-dimensional representations if random projections are used. The principal curvature in the expected lower-dimensional representation is proportional to the mean curvature in the original loss space. Hence, the mean curvature in the original loss space determines if saddle points appear, on average, as either minima, maxima, or almost flat regions. We use the connection between expected curvature in random projections and mean curvature in the original space (i.e., the normalized Hessian trace) to compute Hutchinson-type trace estimates without calculating Hessian-vector products as in the original Hutchinson method. Because random projections are not suitable to correctly identify saddle information, we propose to study projections along dominant Hessian directions that are associated with the largest and smallest principal curvatures. We connect our findings to the ongoing debate on loss landscape flatness and generalizability. Finally, for different common image classifiers and a function approximator, we show and compare random and Hessian projections of loss landscapes with up to about $7\times 10^6$ parameters.
Authors: Sarah Pratt, Ian Covert, Rosanne Liu, Ali Farhadi
Open-vocabulary models are a promising new paradigm for image classification. Unlike traditional classification models, open-vocabulary models classify among any arbitrary set of categories specified with natural language during inference. This natural language, called "prompts", typically consists of a set of hand-written templates (e.g., "a photo of a {}") which are completed with each of the category names. This work introduces a simple method to generate higher accuracy prompts, without relying on any explicit knowledge of the task domain and with far fewer hand-constructed sentences. To achieve this, we combine open-vocabulary models with large language models (LLMs) to create Customized Prompts via Language models (CuPL, pronounced "couple"). In particular, we leverage the knowledge contained in LLMs in order to generate many descriptive sentences that contain important discriminating characteristics of the image categories. This allows the model to place a greater importance on these regions in the image when making predictions. We find that this straightforward and general approach improves accuracy on a range of zero-shot image classification benchmarks, including over one percentage point gain on ImageNet. Finally, this simple baseline requires no additional training and remains completely zero-shot. Code available at https://github.com/sarahpratt/CuPL.
Authors: Wanqin Ma, Huifeng Yao, Yiqun Lin, Jiarong Guo, Xiaomeng Li
Developing a deep learning method for medical segmentation tasks heavily relies on a large amount of labeled data. However, the annotations require professional knowledge and are limited in number. Recently, semi-supervised learning has demonstrated great potential in medical segmentation tasks. Most existing methods related to cardiac magnetic resonance images only focus on regular images with similar domains and high image quality. A semi-supervised domain generalization method was developed in [2], which enhances the quality of pseudo labels on varied datasets. In this paper, we follow the strategy in [2] and present a domain generalization method for semi-supervised medical segmentation. Our main goal is to improve the quality of pseudo labels under extreme MRI Analysis with various domains. We perform Fourier transformation on input images to learn low-level statistics and cross-domain information. Then we feed the augmented images as input to the double cross pseudo supervision networks to calculate the variance among pseudo labels. We evaluate our method on the CMRxMotion dataset [1]. With only partially labeled data and without domain labels, our approach consistently generates accurate segmentation results of cardiac magnetic resonance images with different respiratory motions. Code is available at: https://github.com/MAWanqin2002/STACOM2022Ma
Authors: Amin Ghiasi, Ali Shafahi, Reza Ardekani
We propose adaptive weight decay, which automatically tunes the hyper-parameter for weight decay during each training iteration. For classification problems, we propose changing the value of the weight decay hyper-parameter on the fly based on the strength of updates from the classification loss (i.e., gradient of cross-entropy), and the regularization loss (i.e., $\ell_2$-norm of the weights). We show that this simple modification can result in large improvements in adversarial robustness -- an area which suffers from robust overfitting -- without requiring extra data across various datasets and architecture choices. For example, our reformulation results in $20\%$ relative robustness improvement for CIFAR-100, and $10\%$ relative robustness improvement on CIFAR-10 comparing to the best tuned hyper-parameters of traditional weight decay resulting in models that have comparable performance to SOTA robustness methods. In addition, this method has other desirable properties, such as less sensitivity to learning rate, and smaller weight norms, which the latter contributes to robustness to overfitting to label noise, and pruning.
Authors: Igor Zingman, Birgit Stierstorfer, Charlotte Lempp, Fabian Heinemann
We present a system for anomaly detection in histopathological images. In histology, normal samples are usually abundant, whereas anomalous (pathological) cases are scarce or not available. Under such settings, one-class classifiers trained on healthy data can detect out-of-distribution anomalous samples. Such approaches combined with pre-trained Convolutional Neural Network (CNN) representations of images were previously employed for anomaly detection (AD). However, pre-trained off-the-shelf CNN representations may not be sensitive to abnormal conditions in tissues, while natural variations of healthy tissue may result in distant representations. To adapt representations to relevant details in healthy tissue we propose training a CNN on an auxiliary task that discriminates healthy tissue of different species, organs, and staining reagents. Almost no additional labeling workload is required, since healthy samples come automatically with aforementioned labels. During training we enforce compact image representations with a center-loss term, which further improves representations for AD. The proposed system outperforms established AD methods on a published dataset of liver anomalies. Moreover, it provided comparable results to conventional methods specifically tailored for quantification of liver anomalies. We show that our approach can be used for toxicity assessment of candidate drugs at early development stages and thereby may reduce expensive late-stage drug attrition.
Authors: Chen Qian, Yuncheng Gao, Mingyang Han, Zi Wang, Dan Ruan, Yu Shen, Yiping Wu, Yirong Zhou, Chengyan Wang, Boyu Jiang, Ran Tao, Zhigang Wu, Jiazheng Wang, Liuhong Zhu, Yi Guo, Taishan Kang, Jianzhong Lin, Tao Gong, Chen Yang, Guoqiang Fei, Meijin Lin, Di Guo, Jianjun Zhou, Meiyun Wang, Xiaobo Qu
Diffusion magnetic resonance imaging (MRI) is the only imaging modality for non-invasive movement detection of in vivo water molecules, with significant clinical and research applications. Diffusion MRI (DWI) acquired by multi-shot techniques can achieve higher resolution, better signal-to-noise ratio, and lower geometric distortion than single-shot, but suffers from inter-shot motion-induced artifacts. These artifacts cannot be removed prospectively, leading to the absence of artifact-free training labels. Thus, the potential of deep learning in multi-shot DWI reconstruction remains largely untapped. To break the training data bottleneck, here, we propose a Physics-Informed Deep DWI reconstruction method (PIDD) to synthesize high-quality paired training data by leveraging the physical diffusion model (magnitude synthesis) and inter-shot motion-induced phase model (motion phase synthesis). The network is trained only once with 100,000 synthetic samples, achieving encouraging results on multiple realistic in vivo data reconstructions. Advantages over conventional methods include: (a) Better motion artifact suppression and reconstruction stability; (b) Outstanding generalization to multi-scenario reconstructions, including multi-resolution, multi-b-value, multi-undersampling, multi-vendor, and multi-center; (c) Excellent clinical adaptability to patients with verifications by seven experienced doctors (p<0.001). In conclusion, PIDD presents a novel deep learning framework by exploiting the power of MRI physics, providing a cost-effective and explainable way to break the data bottleneck in deep learning medical imaging.
Authors: Weihao Yu, Chenyang Si, Pan Zhou, Mi Luo, Yichen Zhou, Jiashi Feng, Shuicheng Yan, Xinchao Wang
MetaFormer, the abstracted architecture of Transformer, has been found to play a significant role in achieving competitive performance. In this paper, we further explore the capacity of MetaFormer, again, without focusing on token mixer design: we introduce several baseline models under MetaFormer using the most basic or common mixers, and summarize our observations as follows: (1) MetaFormer ensures solid lower bound of performance. By merely adopting identity mapping as the token mixer, the MetaFormer model, termed IdentityFormer, achieves >80% accuracy on ImageNet-1K. (2) MetaFormer works well with arbitrary token mixers. When specifying the token mixer as even a random matrix to mix tokens, the resulting model RandFormer yields an accuracy of >81%, outperforming IdentityFormer. Rest assured of MetaFormer's results when new token mixers are adopted. (3) MetaFormer effortlessly offers state-of-the-art results. With just conventional token mixers dated back five years ago, the models instantiated from MetaFormer already beat state of the art. (a) ConvFormer outperforms ConvNeXt. Taking the common depthwise separable convolutions as the token mixer, the model termed ConvFormer, which can be regarded as pure CNNs, outperforms the strong CNN model ConvNeXt. (b) CAFormer sets new record on ImageNet-1K. By simply applying depthwise separable convolutions as token mixer in the bottom stages and vanilla self-attention in the top stages, the resulting model CAFormer sets a new record on ImageNet-1K: it achieves an accuracy of 85.5% at 224x224 resolution, under normal supervised training without external data or distillation. In our expedition to probe MetaFormer, we also find that a new activation, StarReLU, reduces 71% FLOPs of activation compared with GELU yet achieves better performance. We expect StarReLU to find great potential in MetaFormer-like models alongside other neural networks.
Authors: Zhongxiang Zhou, Yifei Yang, Yue Wang, Rong Xiong
Detecting both known and unknown objects is a fundamental skill for robot manipulation in unstructured environments. Open-set object detection (OSOD) is a promising direction to handle the problem consisting of two subtasks: objects and background separation, and open-set object classification. In this paper, we present Openset RCNN to address the challenging OSOD. To disambiguate unknown objects and background in the first subtask, we propose to use classification-free region proposal network (CF-RPN) which estimates the objectness score of each region purely using cues from object's location and shape preventing overfitting to the training categories. To identify unknown objects in the second subtask, we propose to represent them using the complementary region of known categories in a latent space which is accomplished by a prototype learning network (PLN). PLN performs instance-level contrastive learning to encode proposals to a latent space and builds a compact region centering with a prototype for each known category. Further, we note that the detection performance of unknown objects can not be unbiasedly evaluated on the situation that commonly used object detection datasets are not fully annotated. Thus, a new benchmark is introduced by reorganizing GraspNet-1billion, a robotic grasp pose detection dataset with complete annotation. Extensive experiments demonstrate the merits of our method. We finally show that our Openset RCNN can endow the robot with an open-set perception ability to support robotic rearrangement tasks in cluttered environments. More details can be found in https://sites.google.com/view/openset-rcnn/
Authors: Matheus Viana da Silva, Natália de Carvalho Santos, Julie Ouellette, Baptiste Lacoste, Cesar Henrique Comin
Creating a dataset for training supervised machine learning algorithms can be a demanding task. This is especially true for medical image segmentation since one or more specialists are usually required for image annotation, and creating ground truth labels for just a single image can take up to several hours. In addition, it is paramount that the annotated samples represent well the different conditions that might affect the imaged tissues as well as possible changes in the image acquisition process. This can only be achieved by considering samples that are typical in the dataset as well as atypical, or even outlier, samples. We introduce a new sampling methodology for selecting relevant images from a large dataset in a way that evenly considers both prototypical as well as atypical samples. The methodology involves the generation of a uniform grid from a feature space representing the samples, which is then used for randomly drawing relevant images. The selected images provide a uniform covering of the original dataset, and thus define a heterogeneous set of images that can be annotated and used for training supervised segmentation algorithms. We provide a case example by creating a dataset containing a representative set of blood vessel microscopy images selected from a larger dataset containing thousands of images. The dataset, which we call VessMAP, is being made available online to aid the development of new blood vessel segmentation algorithms.
Authors: Yuanyuan Peng, Lin Pan, Pengpeng Luan, Hongbin Tu, Xiong Li
Automatic segmentation of curvilinear objects in medical images plays an important role in the diagnosis and evaluation of human diseases, yet it is a challenging uncertainty in the complex segmentation tasks due to different issues such as various image appearances, low contrast between curvilinear objects and their surrounding backgrounds, thin and uneven curvilinear structures, and improper background illumination conditions. To overcome these challenges, we present a unique curvilinear structure segmentation framework based on an oriented derivative of stick (ODoS) filter and a deep learning network for curvilinear object segmentation in medical images. Currently, a large number of deep learning models emphasize developing deep architectures and ignore capturing the structural features of curvilinear objects, which may lead to unsatisfactory results. Consequently, a new approach that incorporates an ODoS filter as part of a deep learning network is presented to improve the spatial attention of curvilinear objects. Specifically, the input image is transfered into four-channel image constructed by the ODoS filter. In which, the original image is considered the principal part to describe various image appearance and complex background illumination conditions, a multi-step strategy is used to enhance the contrast between curvilinear objects and their surrounding backgrounds, and a vector field is applied to discriminate thin and uneven curvilinear structures. Subsequently, a deep learning framework is employed to extract various structural features for curvilinear object segmentation in medical images. The performance of the computational model is validated in experiments conducted on the publicly available DRIVE, STARE and CHASEDB1 datasets. The experimental results indicate that the presented model yields surprising results compared with those of some state-of-the-art methods.
Authors: Joao P. C. Bertoldo, Santiago Velasco-Forero, Jesus Angulo, Etienne Decencière
We propose an incremental improvement to Fully Convolutional Data Description (FCDD), an adaptation of the one-class classification approach from anomaly detection to image anomaly segmentation (a.k.a. anomaly localization). We analyze its original loss function and propose a substitute that better resembles its predecessor, the Hypersphere Classifier (HSC). Both are compared on the MVTec Anomaly Detection Dataset (MVTec-AD) -- training images are flawless objects/textures and the goal is to segment unseen defects -- showing that consistent improvement is achieved by better designing the pixel-wise supervision.
Authors: Hao Liang, Josue Ortega Caro, Vikram Maheshri, Ankit B. Patel, Guha Balakrishnan
Training dataset biases are by far the most scrutinized factors when explaining algorithmic biases of neural networks. In contrast, hyperparameters related to the neural network architecture have largely been ignored even though different network parameterizations are known to induce different implicit biases over learned features. For example, convolutional kernel size is known to affect the frequency content of features learned in CNNs. In this work, we present a causal framework for linking an architectural hyperparameter to out-of-distribution algorithmic bias. Our framework is experimental, in that we train several versions of a network with an intervention to a specific hyperparameter, and measure the resulting causal effect of this choice on performance bias when a particular out-of-distribution image perturbation is applied. In our experiments, we focused on measuring the causal relationship between convolutional kernel size and face analysis classification bias across different subpopulations (race/gender), with respect to high-frequency image details. We show that modifying kernel size, even in one layer of a CNN, changes the frequency content of learned features significantly across data subgroups leading to biased generalization performance even in the presence of a balanced dataset.
Authors: Marco Jiralerspong, Avishek Joey Bose, Ian Gemp, Chongli Qin, Yoram Bachrach, Gauthier Gidel
The past few years have seen impressive progress in the development of deep generative models capable of producing high-dimensional, complex, and photo-realistic data. However, current methods for evaluating such models remain incomplete: standard likelihood-based metrics do not always apply and rarely correlate with perceptual fidelity, while sample-based metrics, such as FID, are insensitive to overfitting, i.e., inability to generalize beyond the training set. To address these limitations, we propose a new metric called the Feature Likelihood Score (FLS), a parametric sample-based score that uses density estimation to provide a comprehensive trichotomic evaluation accounting for novelty (i.e., different from the training samples), fidelity, and diversity of generated samples. We empirically demonstrate the ability of FLS to identify specific overfitting problem cases, where previously proposed metrics fail. We also extensively evaluate FLS on various image datasets and model classes, demonstrating its ability to match intuitions of previous metrics like FID while offering a more comprehensive evaluation of generative models. Code is available at https://github.com/marcojira/fls.
Authors: Thomas Yerxa, Yilun Kuang, Eero Simoncelli, SueYeon Chung
The efficient coding hypothesis proposes that the response properties of sensory systems are adapted to the statistics of their inputs such that they capture maximal information about the environment, subject to biological constraints. While elegant, information theoretic properties are notoriously difficult to measure in practical settings or to employ as objective functions in optimization. This difficulty has necessitated that computational models designed to test the hypothesis employ several different information metrics ranging from approximations and lower bounds to proxy measures like reconstruction error. Recent theoretical advances have characterized a novel and ecologically relevant efficiency metric, the manifold capacity, which is the number of object categories that may be represented in a linearly separable fashion. However, calculating manifold capacity is a computationally intensive iterative procedure that until now has precluded its use as an objective. Here we outline the simplifying assumptions that allow manifold capacity to be optimized directly, yielding Maximum Manifold Capacity Representations (MMCR). The resulting method is closely related to and inspired by advances in the field of self supervised learning (SSL), and we demonstrate that MMCRs are competitive with state of the art results on standard SSL benchmarks. Empirical analyses reveal differences between MMCRs and representations learned by other SSL frameworks, and suggest a mechanism by which manifold compression gives rise to class separability. Finally we evaluate a set of SSL methods on a suite of neural predictivity benchmarks, and find MMCRs are higly competitive as models of the ventral stream.
Authors: Wenxiao Wang, Wei Chen, Qibo Qiu, Long Chen, Boxi Wu, Binbin Lin, Xiaofei He, Wei Liu
While features of different scales are perceptually important to visual inputs, existing vision transformers do not yet take advantage of them explicitly. To this end, we first propose a cross-scale vision transformer, CrossFormer. It introduces a cross-scale embedding layer (CEL) and a long-short distance attention (LSDA). On the one hand, CEL blends each token with multiple patches of different scales, providing the self-attention module itself with cross-scale features. On the other hand, LSDA splits the self-attention module into a short-distance one and a long-distance counterpart, which not only reduces the computational burden but also keeps both small-scale and large-scale features in the tokens. Moreover, through experiments on CrossFormer, we observe another two issues that affect vision transformers' performance, i.e., the enlarging self-attention maps and amplitude explosion. Thus, we further propose a progressive group size (PGS) paradigm and an amplitude cooling layer (ACL) to alleviate the two issues, respectively. The CrossFormer incorporating with PGS and ACL is called CrossFormer++. Extensive experiments show that CrossFormer++ outperforms the other vision transformers on image classification, object detection, instance segmentation, and semantic segmentation tasks. The code will be available at: https://github.com/cheerss/CrossFormer.
Authors: Pingping Cai, Zhenyao Wu, Xinyi Wu, Song Wang
Designing a point cloud upsampler, which aims to generate a clean and dense point cloud given a sparse point representation, is a fundamental and challenging problem in computer vision. A line of attempts achieves this goal by establishing a point-to-point mapping function via deep neural networks. However, these approaches are prone to produce outlier points due to the lack of explicit surface-level constraints. To solve this problem, we introduce a novel surface regularizer into the upsampler network by forcing the neural network to learn the underlying parametric surface represented by bicubic functions and rotation functions, where the new generated points are then constrained on the underlying surface. These designs are integrated into two different networks for two tasks that take advantages of upsampling layers - point cloud upsampling and point cloud completion for evaluation. The state-of-the-art experimental results on both tasks demonstrate the effectiveness of the proposed method. The code is available at https://github.com/corecai163/PSCU.
Authors: Minjong Lee, Dongwoo Kim
We question the current evaluation practice on diffusion-based purification methods. Diffusion-based purification methods aim to remove adversarial effects from an input data point at test time. The approach gains increasing attention as an alternative to adversarial training due to the disentangling between training and testing. Well-known white-box attacks are often employed to measure the robustness of the purification. However, it is unknown whether these attacks are the most effective for the diffusion-based purification since the attacks are often tailored for adversarial training. We analyze the current practices and provide a new guideline for measuring the robustness of purification methods against adversarial attacks. Based on our analysis, we further propose a new purification strategy improving robustness compared to the current diffusion-based purification methods.
Authors: Cheng Jin, Zhengrui Guo, Yi Lin, Luyang Luo, Hao Chen
Deep learning has seen rapid growth in recent years and achieved state-of-the-art performance in a wide range of applications. However, training models typically requires expensive and time-consuming collection of large quantities of labeled data. This is particularly true within the scope of medical imaging analysis (MIA), where data are limited and labels are expensive to be acquired. Thus, label-efficient deep learning methods are developed to make comprehensive use of the labeled data as well as the abundance of unlabeled and weak-labeled data. In this survey, we extensively investigated over 300 recent papers to provide a comprehensive overview of recent progress on label-efficient learning strategies in MIA. We first present the background of label-efficient learning and categorize the approaches into different schemes. Next, we examine the current state-of-the-art methods in detail through each scheme. Specifically, we provide an in-depth investigation, covering not only canonical semi-supervised, self-supervised, and multi-instance learning schemes, but also recently emerged active and annotation-efficient learning strategies. Moreover, as a comprehensive contribution to the field, this survey not only elucidates the commonalities and unique features of the surveyed methods but also presents a detailed analysis of the current challenges in the field and suggests potential avenues for future research.
Authors: Ce Zheng, Xianpeng Liu, Mengyuan Liu, Tianfu Wu, Guo-Jun Qi, Chen Chen
Human mesh recovery (HMR) provides rich human body information for various real-world applications. While image-based HMR methods have achieved impressive results, they often struggle to recover humans in dynamic scenarios, leading to temporal inconsistencies and non-smooth 3D motion predictions due to the absence of human motion. In contrast, video-based approaches leverage temporal information to mitigate this issue. In this paper, we present DiffMesh, an innovative motion-aware Diffusion-like framework for video-based HMR. DiffMesh establishes a bridge between diffusion models and human motion, efficiently generating accurate and smooth output mesh sequences by incorporating human motion within the forward process and reverse process in the diffusion model. Extensive experiments are conducted on the widely used datasets (Human3.6M \cite{h36m_pami} and 3DPW \cite{pw3d2018}), which demonstrate the effectiveness and efficiency of our DiffMesh. Visual comparisons in real-world scenarios further highlight DiffMesh's suitability for practical applications.
Authors: Mehmet Aygün, Oisin Mac Aodha
We introduce SAOR, a novel approach for estimating the 3D shape, texture, and viewpoint of an articulated object from a single image captured in the wild. Unlike prior approaches that rely on pre-defined category-specific 3D templates or tailored 3D skeletons, SAOR learns to articulate shapes from single-view image collections with a skeleton-free part-based model without requiring any 3D object shape priors. To prevent ill-posed solutions, we propose a cross-instance consistency loss that exploits disentangled object shape deformation and articulation. This is helped by a new silhouette-based sampling mechanism to enhance viewpoint diversity during training. Our method only requires estimated object silhouettes and relative depth maps from off-the-shelf pre-trained networks during training. At inference time, given a single-view image, it efficiently outputs an explicit mesh representation. We obtain improved qualitative and quantitative results on challenging quadruped animals compared to relevant existing work.
Authors: Haotian Bai, Yuanhuiyi Lyu, Lutao Jiang, Sijia Li, Haonan Lu, Xiaodong Lin, Lin Wang
Recent advances have shown promise in merging neural radiance fields (NeRFs) with pre-trained diffusion models for text-to-3D object generation. However, one enduring challenge is their inadequate capability to accurately parse and regenerate consistent multi-object environments. Specifically, these models encounter difficulties in accurately representing quantity and style prompted by multi-object texts, often resulting in a collapse of the rendering fidelity that fails to match the semantic intricacies. Moreover, amalgamating these elements into a coherent 3D scene is a substantial challenge, stemming from generic distribution inherent in diffusion models. To tackle the issue of 'guidance collapse' and enhance consistency, we propose a novel framework, dubbed CompoNeRF, by integrating an editable 3D scene layout with object specific and scene-wide guidance mechanisms. It initiates by interpreting a complex text into an editable 3D layout populated with multiple NeRFs, each paired with a corresponding subtext prompt for precise object depiction. Next, a tailored composition module seamlessly blends these NeRFs, promoting consistency, while the dual-level text guidance reduces ambiguity and boosts accuracy. Noticeably, the unique modularity of CompoNeRF permits NeRF decomposition. This enables flexible scene editing and recomposition into new scenes based on the edited layout or text prompts. Utilizing the open source Stable Diffusion model, CompoNeRF not only generates scenes with high fidelity but also paves the way for innovative multi-object composition using editable 3D layouts. Remarkably, our framework achieves up to a 54\% improvement in performance, as measured by the multi-view CLIP score metric. Code is available at https://github.com/hbai98/Componerf.
Authors: Yongliang Shen, Kaitao Song, Xu Tan, Dongsheng Li, Weiming Lu, Yueting Zhuang
Solving complicated AI tasks with different domains and modalities is a key step toward artificial general intelligence. While there are numerous AI models available for various domains and modalities, they cannot handle complicated AI tasks autonomously. Considering large language models (LLMs) have exhibited exceptional abilities in language understanding, generation, interaction, and reasoning, we advocate that LLMs could act as a controller to manage existing AI models to solve complicated AI tasks, with language serving as a generic interface to empower this. Based on this philosophy, we present HuggingGPT, an LLM-powered agent that leverages LLMs (e.g., ChatGPT) to connect various AI models in machine learning communities (e.g., Hugging Face) to solve AI tasks. Specifically, we use ChatGPT to conduct task planning when receiving a user request, select models according to their function descriptions available in Hugging Face, execute each subtask with the selected AI model, and summarize the response according to the execution results. By leveraging the strong language capability of ChatGPT and abundant AI models in Hugging Face, HuggingGPT can tackle a wide range of sophisticated AI tasks spanning different modalities and domains and achieve impressive results in language, vision, speech, and other challenging tasks, which paves a new way towards the realization of artificial general intelligence.
Authors: Yuang Ai, Xiaoqiang Zhou, Huaibo Huang, Lei Zhang, Ran He
Unsupervised Domain Adaptation (UDA) can effectively address domain gap issues in real-world image Super-Resolution (SR) by accessing both the source and target data. Considering privacy policies or transmission restrictions of source data in practical scenarios, we propose a SOurce-free Domain Adaptation framework for image SR (SODA-SR) to address this issue, i.e., adapt a source-trained model to a target domain with only unlabeled target data. SODA-SR leverages the source-trained model to generate refined pseudo-labels for teacher-student learning. To better utilize pseudo-labels, we propose a novel wavelet-based augmentation method, named Wavelet Augmentation Transformer (WAT), which can be flexibly incorporated with existing networks, to implicitly produce useful augmented data. WAT learns low-frequency information of varying levels across diverse samples, which is aggregated efficiently via deformable attention. Furthermore, an uncertainty-aware self-training mechanism is proposed to improve the accuracy of pseudo-labels, with inaccurate predictions being rectified by uncertainty estimation. To acquire better SR results and avoid overfitting pseudo-labels, several regularization losses are proposed to constrain target LR and SR images in the frequency domain. Experiments show that without accessing source data, SODA-SR outperforms state-of-the-art UDA methods in both synthetic$\rightarrow$real and real$\rightarrow$real adaptation settings, and is not constrained by specific network architectures.
Authors: Rahul Chand, Rajat Arora, K Ram Prabhakar, R Venkatesh Babu
We present a framework to use recently introduced Capsule Networks for solving the problem of Optical Flow, one of the fundamental computer vision tasks. Most of the existing state of the art deep architectures either uses a correlation oepration to match features from them. While correlation layer is sensitive to the choice of hyperparameters and does not put a prior on the underlying structure of the object, spatio temporal features will be limited by the network's receptive field. Also, we as humans look at moving objects as whole, something which cannot be encoded by correlation or spatio temporal features. Capsules, on the other hand, are specialized to model seperate entities and their pose as a continuous matrix. Thus, we show that a simpler linear operation over poses of the objects detected by the capsules in enough to model flow. We show reslts on a small toy dataset where we outperform FlowNetC and PWC-Net models.
Authors: Andrei-Timotei Ardelean, Tim Weyrich
We propose a novel method for Zero-Shot Anomaly Localization on textures. The task refers to identifying abnormal regions in an otherwise homogeneous image. To obtain a high-fidelity localization, we leverage a bijective mapping derived from the 1-dimensional Wasserstein Distance. As opposed to using holistic distances between distributions, the proposed approach allows pinpointing the non-conformity of a pixel in a local context with increased precision. By aggregating the contribution of the pixel to the errors of all nearby patches we obtain a reliable anomaly score estimate. We validate our solution on several datasets and obtain more than a 40% reduction in error over the previous state of the art on the MVTec AD dataset in a zero-shot setting. Also see https://reality.tf.fau.de/pub/ardelean2024highfidelity.html.
Authors: Tao Zhuo, Zhiyong Cheng, Zan Gao, Hehe Fan, Mohan Kankanhalli
Continual Learning (CL) aims at incrementally learning new tasks without forgetting the knowledge acquired from old ones. Experience Replay (ER) is a simple and effective rehearsal-based strategy, which optimizes the model with current training data and a subset of old samples stored in a memory buffer. To further reduce forgetting, recent approaches extend ER with various techniques, such as model regularization and memory sampling. However, the prediction consistency between the new model and the old one on current training data has been seldom explored, resulting in less knowledge preserved when few previous samples are available. To address this issue, we propose a CL method with Strong Experience Replay (SER), which additionally utilizes future experiences mimicked on the current training data, besides distilling past experience from the memory buffer. In our method, the updated model will produce approximate outputs as its original ones, which can effectively preserve the acquired knowledge. Experimental results on multiple image classification datasets show that our SER method surpasses the state-of-the-art methods by a noticeable margin.
Authors: Tao Huang, Yuan Zhang, Mingkai Zheng, Shan You, Fei Wang, Chen Qian, Chang Xu
The representation gap between teacher and student is an emerging topic in knowledge distillation (KD). To reduce the gap and improve the performance, current methods often resort to complicated training schemes, loss functions, and feature alignments, which are task-specific and feature-specific. In this paper, we state that the essence of these methods is to discard the noisy information and distill the valuable information in the feature, and propose a novel KD method dubbed DiffKD, to explicitly denoise and match features using diffusion models. Our approach is based on the observation that student features typically contain more noises than teacher features due to the smaller capacity of student model. To address this, we propose to denoise student features using a diffusion model trained by teacher features. This allows us to perform better distillation between the refined clean feature and teacher feature. Additionally, we introduce a light-weight diffusion model with a linear autoencoder to reduce the computation cost and an adaptive noise matching module to improve the denoising performance. Extensive experiments demonstrate that DiffKD is effective across various types of features and achieves state-of-the-art performance consistently on image classification, object detection, and semantic segmentation tasks. Code is available at https://github.com/hunto/DiffKD.
Authors: Aristotelis Ballas, Christos Diou
In search of robust and generalizable machine learning models, Domain Generalization (DG) has gained significant traction during the past few years. The goal in DG is to produce models which continue to perform well when presented with data distributions different from the ones available during training. While deep convolutional neural networks (CNN) have been able to achieve outstanding performance on downstream computer vision tasks, they still often fail to generalize on previously unseen data Domains. Therefore, in this work we focus on producing a model which is able to remain robust under data distribution shift and propose an alternative regularization technique for convolutional neural network architectures in the single-source DG image classification setting. To mitigate the problem caused by domain shift between source and target data, we propose augmenting intermediate feature maps of CNNs. Specifically, we pass them through a novel Augmentation Layer} to prevent models from overfitting on the training set and improve their cross-domain generalization. To the best of our knowledge, this is the first paper proposing such a setup for the DG image classification setting. Experiments on the DG benchmark datasets of PACS, VLCS, Office-Home and TerraIncognita validate the effectiveness of our method, in which our model surpasses state-of-the-art algorithms in most cases.
Authors: Jian Chen, Ruiyi Zhang, Tong Yu, Rohan Sharma, Zhiqiang Xu, Tong Sun, Changyou Chen
Learning from noisy labels is an important and long-standing problem in machine learning for real applications. One of the main research lines focuses on learning a label corrector to purify potential noisy labels. However, these methods typically rely on strict assumptions and are limited to certain types of label noise. In this paper, we reformulate the label-noise problem from a generative-model perspective, $\textit{i.e.}$, labels are generated by gradually refining an initial random guess. This new perspective immediately enables existing powerful diffusion models to seamlessly learn the stochastic generative process. Once the generative uncertainty is modeled, we can perform classification inference using maximum likelihood estimation of labels. To mitigate the impact of noisy labels, we propose the $\textbf{L}$abel-$\textbf{R}$etrieval-$\textbf{A}$ugmented (LRA) diffusion model, which leverages neighbor consistency to effectively construct pseudo-clean labels for diffusion training. Our model is flexible and general, allowing easy incorporation of different types of conditional information, $\textit{e.g.}$, use of pre-trained models, to further boost model performance. Extensive experiments are conducted for evaluation. Our model achieves new state-of-the-art (SOTA) results on all the standard real-world benchmark datasets. Remarkably, by incorporating conditional information from the powerful CLIP model, our method can boost the current SOTA accuracy by 10-20 absolute points in many cases.
Authors: Hao Zhang, Yanbo Xu, Tianyuan Dai, Yu-Wing Tai, Chi-Keung Tang
The ability to create high-quality 3D faces from a single image has become increasingly important with wide applications in video conferencing, AR/VR, and advanced video editing in movie industries. In this paper, we propose Face Diffusion NeRF (FaceDNeRF), a new generative method to reconstruct high-quality Face NeRFs from single images, complete with semantic editing and relighting capabilities. FaceDNeRF utilizes high-resolution 3D GAN inversion and expertly trained 2D latent-diffusion model, allowing users to manipulate and construct Face NeRFs in zero-shot learning without the need for explicit 3D data. With carefully designed illumination and identity preserving loss, as well as multi-modal pre-training, FaceDNeRF offers users unparalleled control over the editing process enabling them to create and edit face NeRFs using just single-view images, text prompts, and explicit target lighting. The advanced features of FaceDNeRF have been designed to produce more impressive results than existing 2D editing approaches that rely on 2D segmentation maps for editable attributes. Experiments show that our FaceDNeRF achieves exceptionally realistic results and unprecedented flexibility in editing compared with state-of-the-art 3D face reconstruction and editing methods. Our code will be available at https://github.com/BillyXYB/FaceDNeRF.
Authors: Thomas Walker, Octave Mariotti, Amir Vaxman, Hakan Bilen
We introduce Explicit Neural Surfaces (ENS), an efficient surface reconstruction method that learns an explicitly defined continuous surface from multiple views. We use a series of neural deformation fields to progressively transform a continuous input surface to a target shape. By sampling meshes as discrete surface proxies, we train the deformation fields through efficient differentiable rasterization, and attain a mesh-independent and smooth surface representation. By using Laplace-Beltrami eigenfunctions as an intrinsic positional encoding alongside standard extrinsic Fourier features, our approach can capture fine surface details. ENS trains 1 to 2 orders of magnitude faster and can extract meshes of higher quality compared to implicit representations, whilst maintaining competitive surface reconstruction performance and real-time capabilities. Finally, we apply our approach to learn a collection of objects in a single model, and achieve disentangled interpolations between different shapes, their surface details, and textures.
Authors: Tan H. Nguyen, Dinkar Juyal, Jin Li, Aaditya Prakash, Shima Nofallah, Chintan Shah, Sai Chowdary Gullapally, Limin Yu, Michael Griffin, Anand Sampat, John Abel, Justin Lee, Amaro Taylor-Weiner
Domain generalization is critical for real-world applications of machine learning to microscopy images, including histopathology and fluorescence imaging. Artifacts in these modalities arise through a complex combination of factors relating to tissue collection and laboratory processing, as well as factors intrinsic to patient samples. In fluorescence imaging, these artifacts stem from variations across experimental batches. The complexity and subtlety of these artifacts make the enumeration of data domains intractable. Therefore, augmentation-based methods of domain generalization that require domain identifiers and manual fine-tuning are inadequate in this setting. To overcome this challenge, we introduce ContriMix, a domain generalization technique that learns to generate synthetic images by disentangling and permuting the biological content ("content") and technical variations ("attributes") in microscopy images. ContriMix does not rely on domain identifiers or handcrafted augmentations and makes no assumptions about the input characteristics of images. We assess the performance of ContriMix on two pathology datasets dealing with patch classification and Whole Slide Image label prediction tasks respectively (Camelyon17-WILDS and RCC subtyping), and one fluorescence microscopy dataset (RxRx1-WILDS). Without any access to domain identifiers at train or test time, ContriMix performs similar or better than current state-of-the-art methods in all these datasets, motivating its usage for microscopy image analysis in real-world settings where domain information is hard to come by. The code for ContriMix can be found at https://gitlab.com/huutan86/contrimix
Authors: Yash Bhalgat, Iro Laina, João F. Henriques, Andrew Zisserman, Andrea Vedaldi
Instance segmentation in 3D is a challenging task due to the lack of large-scale annotated datasets. In this paper, we show that this task can be addressed effectively by leveraging instead 2D pre-trained models for instance segmentation. We propose a novel approach to lift 2D segments to 3D and fuse them by means of a neural field representation, which encourages multi-view consistency across frames. The core of our approach is a slow-fast clustering objective function, which is scalable and well-suited for scenes with a large number of objects. Unlike previous approaches, our method does not require an upper bound on the number of objects or object tracking across frames. To demonstrate the scalability of the slow-fast clustering, we create a new semi-realistic dataset called the Messy Rooms dataset, which features scenes with up to 500 objects per scene. Our approach outperforms the state-of-the-art on challenging scenes from the ScanNet, Hypersim, and Replica datasets, as well as on our newly created Messy Rooms dataset, demonstrating the effectiveness and scalability of our slow-fast clustering method.
Authors: Frank Permenter, Chenyang Yuan
Denoising is intuitively related to projection. Indeed, under the manifold hypothesis, adding random noise is approximately equivalent to orthogonal perturbation. Hence, learning to denoise is approximately learning to project. In this paper, we use this observation to reinterpret denoising diffusion models as approximate gradient descent applied to the Euclidean distance function. We then provide straight-forward convergence analysis of the DDIM sampler under simple assumptions on the projection-error of the denoiser. Finally, we propose a new sampler based on two simple modifications to DDIM using insights from our theoretical results. In as few as 5-10 function evaluations, our sampler achieves state-of-the-art FID scores on pretrained CIFAR-10 and CelebA models and can generate high quality samples on latent diffusion models.
Authors: Marcin Przewięźlikowski, Mateusz Pyla, Bartosz Zieliński, Bartłomiej Twardowski, Jacek Tabor, Marek Śmieja
Self-supervised learning (SSL) is a powerful technique for learning robust representations from unlabeled data. By learning to remain invariant to applied data augmentations, methods such as SimCLR and MoCo are able to reach quality on par with supervised approaches. However, this invariance may be harmful to solving some downstream tasks which depend on traits affected by augmentations used during pretraining, such as color. In this paper, we propose to foster sensitivity to such characteristics in the representation space by modifying the projector network, a common component of self-supervised architectures. Specifically, we supplement the projector with information about augmentations applied to images. In order for the projector to take advantage of this auxiliary conditioning when solving the SSL task, the feature extractor learns to preserve the augmentation information in its representations. Our approach, coined Conditional Augmentation-aware Self-supervised Learning (CASSLE), is directly applicable to typical joint-embedding SSL methods regardless of their objective functions. Moreover, it does not require major changes in the network architecture or prior knowledge of downstream tasks. In addition to an analysis of sensitivity towards different data augmentations, we conduct a series of experiments, which show that CASSLE improves over various SSL methods, reaching state-of-the-art performance in multiple downstream tasks.
Authors: Seungjoo Shin, Jaesik Park
In this paper, we propose \textit{binary radiance fields} (BiRF), a storage-efficient radiance field representation employing binary feature encoding that encodes local features using binary encoding parameters in a format of either $+1$ or $-1$. This binarization strategy lets us represent the feature grid with highly compact feature encoding and a dramatic reduction in storage size. Furthermore, our 2D-3D hybrid feature grid design enhances the compactness of feature encoding as the 3D grid includes main components while 2D grids capture details. In our experiments, binary radiance field representation successfully outperforms the reconstruction performance of state-of-the-art (SOTA) efficient radiance field models with lower storage allocation. In particular, our model achieves impressive results in static scene reconstruction, with a PSNR of 32.03 dB for Synthetic-NeRF scenes, 34.48 dB for Synthetic-NSVF scenes, 28.20 dB for Tanks and Temples scenes while only utilizing 0.5 MB of storage space, respectively. We hope the proposed binary radiance field representation will make radiance fields more accessible without a storage bottleneck.
Authors: Linfeng Yuan, Miaojing Shi, Zijie Yue, Qijun Chen
Referring video object segmentation (RVOS) aims to segment the target instance referred by a given text expression in a video clip. The text expression normally contains sophisticated description of the instance's appearance, action, and relation with others. It is therefore rather difficult for a RVOS model to capture all these attributes correspondingly in the video; in fact, the model often favours more on the action- and relation-related visual attributes of the instance. This can end up with partial or even incorrect mask prediction of the target instance. We tackle this problem by taking a subject-centric short text expression from the original long text expression. The short one retains only the appearance-related information of the target instance so that we can use it to focus the model's attention on the instance's appearance. We let the model make joint predictions using both long and short text expressions; and insert a long-short cross-attention module to interact the joint features and a long-short predictions intersection loss to regulate the joint predictions. Besides the improvement on the linguistic part, we also introduce a forward-backward visual consistency loss, which utilizes optical flows to warp visual features between the annotated frames and their temporal neighbors for consistency. We build our method on top of two state of the art pipelines. Extensive experiments on A2D-Sentences, Refer-YouTube-VOS, JHMDB-Sentences and Refer-DAVIS17 show impressive improvements of our method.
Authors: Jiafan Zhuang, Duan Yuan, Rihong Yan, Weixin Huang, Wenji Li, Zhun Fan
Distance estimation plays an important role for path planning and collision avoidance of swarm UAVs. However, the lack of annotated data seriously hinders the related studies. In this work, we build and present a UAVDE dataset for UAV distance estimation, in which distance between two UAVs is obtained by UWB sensors. During experiments, we surprisingly observe that the stereo triangulation cannot stand for UAV scenes. The core reason is the position deviation issue due to long shooting distance and camera vibration, which is common in UAV scenes. To tackle this issue, we propose a novel position correction module, which can directly predict the offset between the observed positions and the actual ones and then perform compensation in stereo triangulation calculation. Besides, to further boost performance on hard samples, we propose a dynamic iterative correction mechanism, which is composed of multiple stacked PCMs and a gating mechanism to adaptively determine whether further correction is required according to the difficulty of data samples. We conduct extensive experiments on UAVDE, and our method can achieve a significant performance improvement over a strong baseline (by reducing the relative difference from 49.4% to 9.8%), which demonstrates its effectiveness and superiority. The code and dataset are available at https://github.com/duanyuan13/PCM.
Authors: Zilun Zhang, Tiancheng Zhao, Yulong Guo, Jianwei Yin
Pre-trained Vision-Language Models (VLMs) utilizing extensive image-text paired data have demonstrated unprecedented image-text association capabilities, achieving remarkable results across various downstream tasks. A critical challenge is how to make use of existing large-scale pre-trained VLMs, which are trained on common objects, to perform the domain-specific transfer for accomplishing domain-related downstream tasks. A critical challenge is how to make use of existing large-scale pre-trained VLMs, which are trained on common objects, to perform the domain-specific transfer for accomplishing domain-related downstream tasks. In this paper, we propose a new framework that includes the Domain pre-trained Vision-Language Model (DVLM), bridging the gap between the General Vision-Language Model (GVLM) and domain-specific downstream tasks. Moreover, we present an image-text paired dataset in the field of remote sensing (RS), RS5M, which has 5 million RS images with English descriptions. The dataset is obtained from filtering publicly available image-text paired datasets and captioning label-only RS datasets with pre-trained VLM. These constitute the first large-scale RS image-text paired dataset. Additionally, we fine-tuned the CLIP model and tried several Parameter-Efficient Fine-Tuning methods on RS5M to implement the DVLM. Experimental results show that our proposed dataset is highly effective for various tasks, and our model GeoRSCLIP improves upon the baseline or previous state-of-the-art model by $3\%\sim20\%$ in Zero-shot Classification (ZSC), $3\%\sim6\%$ in Remote Sensing Cross-Modal Text-Image Retrieval (RSCTIR) and $4\%\sim5\%$ in Semantic Localization (SeLo) tasks. Dataset and models have been released in: \url{https://github.com/om-ai-lab/RS5M}.
Authors: Frano Rajič, Lei Ke, Yu-Wing Tai, Chi-Keung Tang, Martin Danelljan, Fisher Yu
The Segment Anything Model (SAM) has established itself as a powerful zero-shot image segmentation model, enabled by efficient point-centric annotation and prompt-based models. While click and brush interactions are both well explored in interactive image segmentation, the existing methods on videos focus on mask annotation and propagation. This paper presents SAM-PT, a novel method for point-centric interactive video segmentation, empowered by SAM and long-term point tracking. SAM-PT leverages robust and sparse point selection and propagation techniques for mask generation. Compared to traditional object-centric mask propagation strategies, we uniquely use point propagation to exploit local structure information agnostic to object semantics. We highlight the merits of point-based tracking through direct evaluation on the zero-shot open-world Unidentified Video Objects (UVO) benchmark. Our experiments on popular video object segmentation and multi-object segmentation tracking benchmarks, including DAVIS, YouTube-VOS, and BDD100K, suggest that a point-based segmentation tracker yields better zero-shot performance and efficient interactions. We release our code that integrates different point trackers and video segmentation benchmarks at https://github.com/SysCV/sam-pt.
Authors: Francesco Cagnetta, Leonardo Petrini, Umberto M. Tomasini, Alessandro Favero, Matthieu Wyart
Deep learning algorithms demonstrate a surprising ability to learn high-dimensional tasks from limited examples. This is commonly attributed to the depth of neural networks, enabling them to build a hierarchy of abstract, low-dimensional data representations. However, how many training examples are required to learn such representations remains unknown. To quantitatively study this question, we introduce the Random Hierarchy Model: a family of synthetic tasks inspired by the hierarchical structure of language and images. The model is a classification task where each class corresponds to a group of high-level features, chosen among several equivalent groups associated with the same class. In turn, each feature corresponds to a group of sub-features chosen among several equivalent ones and so on, following a hierarchy of composition rules. We find that deep networks learn the task by developing internal representations invariant to exchanging equivalent groups. Moreover, the number of data required corresponds to the point where correlations between low-level features and classes become detectable. Overall, our results indicate how deep networks overcome the curse of dimensionality by building invariant representations, and provide an estimate of the number of data required to learn a hierarchical task.
Authors: Liangzhe Yuan, Nitesh Bharadwaj Gundavarapu, Long Zhao, Hao Zhou, Yin Cui, Lu Jiang, Xuan Yang, Menglin Jia, Tobias Weyand, Luke Friedman, Mikhail Sirotenko, Huisheng Wang, Florian Schroff, Hartwig Adam, Ming-Hsuan Yang, Ting Liu, Boqing Gong
We evaluate existing foundation models video understanding capabilities using a carefully designed experiment protocol consisting of three hallmark tasks (action recognition, temporal localization, and spatiotemporal localization), eight datasets well received by the community, and four adaptation methods tailoring a foundation model (FM) for a downstream task. Moreover, we propose a scalar VideoGLUE score (VGS) to measure an FMs efficacy and efficiency when adapting to general video understanding tasks. Our main findings are as follows. First, task-specialized models significantly outperform the six FMs studied in this work, in sharp contrast to what FMs have achieved in natural language and image understanding. Second,video-native FMs, whose pretraining data contains the video modality, are generally better than image-native FMs in classifying motion-rich videos, localizing actions in time, and understanding a video of more than one action. Third, the video-native FMs can perform well on video tasks under light adaptations to downstream tasks(e.g., freezing the FM backbones), while image-native FMs win in full end-to-end finetuning. The first two observations reveal the need and tremendous opportunities to conduct research on video-focused FMs, and the last confirms that both tasks and adaptation methods matter when it comes to the evaluation of FMs. Our code is released under: https://github.com/tensorflow/models/tree/master/official/projects/videoglue.
Authors: Alexander Ziller, Ayhan Can Erdur, Marwa Trigui, Alp Güvenir, Tamara T. Mueller, Philip Müller, Friederike Jungmann, Johannes Brandt, Jan Peeken, Rickmer Braren, Daniel Rueckert, Georgios Kaissis
Training Artificial Intelligence (AI) models on 3D images presents unique challenges compared to the 2D case: Firstly, the demand for computational resources is significantly higher, and secondly, the availability of large datasets for pre-training is often limited, impeding training success. This study proposes a simple approach of adapting 2D networks with an intermediate feature representation for processing 3D images. Our method employs attention pooling to learn to assign each slice an importance weight and, by that, obtain a weighted average of all 2D slices. These weights directly quantify the contribution of each slice to the contribution and thus make the model prediction inspectable. We show on all 3D MedMNIST datasets as benchmark and two real-world datasets consisting of several hundred high-resolution CT or MRI scans that our approach performs on par with existing methods. Furthermore, we compare the in-built interpretability of our approach to HiResCam, a state-of-the-art retrospective interpretability approach.
Authors: Guanlin Li, Kangjie Chen, Yuan Xu, Han Qiu, Tianwei Zhang
Adversarial training is an important topic in robust deep learning, but the community lacks attention to its practical usage. In this paper, we aim to resolve a real-world challenge, i.e., training a model on an imbalanced and noisy dataset to achieve high clean accuracy and adversarial robustness, with our proposed Omnipotent Adversarial Training (OAT) strategy. OAT consists of two innovative methodologies to address the imperfection in the training set. We first introduce an oracle into the adversarial training process to help the model learn a correct data-label conditional distribution. This carefully-designed oracle can provide correct label annotations for adversarial training. We further propose logits adjustment adversarial training to overcome the data imbalance issue, which can help the model learn a Bayes-optimal distribution. Our comprehensive evaluation results show that OAT outperforms other baselines by more than 20% clean accuracy improvement and 10% robust accuracy improvement under complex combinations of data imbalance and label noise scenarios. The code can be found in https://github.com/GuanlinLee/OAT.
Authors: Bowen Xue, Shuang Zhao, Henrik Wann Jensen, Zahra Montazeri
Neural reflectance models are capable of reproducing the spatially-varying appearance of many real-world materials at different scales. Unfortunately, existing techniques such as NeuMIP have difficulties handling materials with strong shadowing effects or detailed specular highlights. In this paper, we introduce a neural appearance model that offers a new level of accuracy. Central to our model is an inception-based core network structure that captures material appearances at multiple scales using parallel-operating kernels and ensures multi-stage features through specialized convolution layers. Furthermore, we encode the inputs into frequency space, introduce a gradient-based loss, and employ it adaptive to the progress of the learning phase. We demonstrate the effectiveness of our method using a variety of synthetic and real examples.
Authors: Guanlin Li, Guowen Xu, Tianwei Zhang
In this paper, we study adversarial training on datasets that obey the long-tailed distribution, which is practical but rarely explored in previous works. Compared with conventional adversarial training on balanced datasets, this process falls into the dilemma of generating uneven adversarial examples (AEs) and an unbalanced feature embedding space, causing the resulting model to exhibit low robustness and accuracy on tail data. To combat that, we theoretically analyze the lower bound of the robust risk to train a model on a long-tailed dataset to obtain the key challenges in addressing the aforementioned dilemmas. Based on it, we propose a new adversarial training framework -- Re-balancing Adversarial Training (REAT). This framework consists of two components: (1) a new training strategy inspired by the effective number to guide the model to generate more balanced and informative AEs; (2) a carefully constructed penalty function to force a satisfactory feature space. Evaluation results on different datasets and model structures prove that REAT can effectively enhance the model's robustness and preserve the model's clean accuracy. The code can be found in https://github.com/GuanlinLee/REAT.
Authors: Enxin Song, Wenhao Chai, Guanhong Wang, Yucheng Zhang, Haoyang Zhou, Feiyang Wu, Haozhe Chi, Xun Guo, Tian Ye, Yanting Zhang, Yan Lu, Jenq-Neng Hwang, Gaoang Wang
Recently, integrating video foundation models and large language models to build a video understanding system can overcome the limitations of specific pre-defined vision tasks. Yet, existing systems can only handle videos with very few frames. For long videos, the computation complexity, memory cost, and long-term temporal connection impose additional challenges. Taking advantage of the Atkinson-Shiffrin memory model, with tokens in Transformers being employed as the carriers of memory in combination with our specially designed memory mechanism, we propose the MovieChat to overcome these challenges. MovieChat achieves state-of-the-art performance in long video understanding, along with the released MovieChat-1K benchmark with 1K long video and 14K manual annotations for validation of the effectiveness of our method.
Authors: Ming Nie, Yujing Xue, Chunwei Wang, Chaoqiang Ye, Hang Xu, Xinge Zhu, Qingqiu Huang, Michael Bi Mi, Xinchao Wang, Li Zhang
Recently, polar-based representation has shown promising properties in perceptual tasks. In addition to Cartesian-based approaches, which separate point clouds unevenly, representing point clouds as polar grids has been recognized as an alternative due to (1) its advantage in robust performance under different resolutions and (2) its superiority in streaming-based approaches. However, state-of-the-art polar-based detection methods inevitably suffer from the feature distortion problem because of the non-uniform division of polar representation, resulting in a non-negligible performance gap compared to Cartesian-based approaches. To tackle this issue, we present PARTNER, a novel 3D object detector in the polar coordinate. PARTNER alleviates the dilemma of feature distortion with global representation re-alignment and facilitates the regression by introducing instance-level geometric information into the detection head. Extensive experiments show overwhelming advantages in streaming-based detection and different resolutions. Furthermore, our method outperforms the previous polar-based works with remarkable margins of 3.68% and 9.15% on Waymo and ONCE validation set, thus achieving competitive results over the state-of-the-art methods.
Authors: Jiashun Suo, Xingzhou Zhang, Weisong Shi, Wei Zhou
Motivated by the advances in deep learning techniques, the application of Unmanned Aerial Vehicle (UAV)-based object detection has proliferated across a range of fields, including vehicle counting, fire detection, and city monitoring. While most existing research studies only a subset of the challenges inherent to UAV-based object detection, there are few studies that balance various aspects to design a practical system for energy consumption reduction. In response, we present the E$^3$-UAV, an edge-based energy-efficient object detection system for UAVs. The system is designed to dynamically support various UAV devices, edge devices, and detection algorithms, with the aim of minimizing energy consumption by deciding the most energy-efficient flight parameters (including flight altitude, flight speed, detection algorithm, and sampling rate) required to fulfill the detection requirements of the task. We first present an effective evaluation metric for actual tasks and construct a transparent energy consumption model based on hundreds of actual flight data to formalize the relationship between energy consumption and flight parameters. Then we present a lightweight energy-efficient priority decision algorithm based on a large quantity of actual flight data to assist the system in deciding flight parameters. Finally, we evaluate the performance of the system, and our experimental results demonstrate that it can significantly decrease energy consumption in real-world scenarios. Additionally, we provide four insights that can assist researchers and engineers in their efforts to study UAV-based object detection further.
Authors: Shijie Huang, Xukun Zhang, Zhiming Cui, He Zhang, Geng Chen, Dinggang Shen
Accurate tissue segmentation of thick-slice fetal brain magnetic resonance (MR) scans is crucial for both reconstruction of isotropic brain MR volumes and the quantification of fetal brain development. However, this task is challenging due to the use of thick-slice scans in clinically-acquired fetal brain data. To address this issue, we propose to leverage high-quality isotropic fetal brain MR volumes (and also their corresponding annotations) as guidance for segmentation of thick-slice scans. Due to existence of significant domain gap between high-quality isotropic volume (i.e., source data) and thick-slice scans (i.e., target data), we employ a domain adaptation technique to achieve the associated knowledge transfer (from high-quality <source> volumes to thick-slice <target> scans). Specifically, we first register the available high-quality isotropic fetal brain MR volumes across different gestational weeks to construct longitudinally-complete source data. To capture domain-invariant information, we then perform Fourier decomposition to extract image content and style codes. Finally, we propose a novel Cycle-Consistent Domain Adaptation Network (C2DA-Net) to efficiently transfer the knowledge learned from high-quality isotropic volumes for accurate tissue segmentation of thick-slice scans. Our C2DA-Net can fully utilize a small set of annotated isotropic volumes to guide tissue segmentation on unannotated thick-slice scans. Extensive experiments on a large-scale dataset of 372 clinically acquired thick-slice MR scans demonstrate that our C2DA-Net achieves much better performance than cutting-edge methods quantitatively and qualitatively.
Authors: Silky Singh, Shripad Deshmukh, Mausoom Sarkar, Balaji Krishnamurthy
Learning object segmentation in image and video datasets without human supervision is a challenging problem. Humans easily identify moving salient objects in videos using the gestalt principle of common fate, which suggests that what moves together belongs together. Building upon this idea, we propose a self-supervised object discovery approach that leverages motion and appearance information to produce high-quality object segmentation masks. Specifically, we redesign the traditional graph cut on images to include motion information in a linear combination with appearance information to produce edge weights. Remarkably, this step produces object segmentation masks comparable to the current state-of-the-art on multiple benchmarks. To further improve performance, we bootstrap a segmentation network trained on these preliminary masks as pseudo-ground truths to learn from its own outputs via self-training. We demonstrate the effectiveness of our approach, named LOCATE, on multiple standard video object segmentation, image saliency detection, and object segmentation benchmarks, achieving results on par with and, in many cases surpassing state-of-the-art methods. We also demonstrate the transferability of our approach to novel domains through a qualitative study on in-the-wild images. Additionally, we present extensive ablation analysis to support our design choices and highlight the contribution of each component of our proposed method.
Authors: Cheng Zhu, Jiayi Zhu, Lijuan Zhang, Xi Wu, Shuqi Yang, Ping Liang, Honghan Chen, Ying Tan
Graph Convolutional Networks (GCNs) can capture non-Euclidean spatial dependence between different brain regions. The graph pooling operator, a crucial element of GCNs, enhances the representation learning capability and facilitates the acquisition of abnormal brain maps. However, most existing research designs graph pooling operators solely from the perspective of nodes while disregarding the original edge features. This confines graph pooling application scenarios and diminishes its ability to capture critical substructures. In this paper, we propose a novel edge-aware hard clustering graph pool (EHCPool), which is tailored to dominant edge features and redefines the clustering process. EHCPool initially introduced the 'Edge-to-Node' score criterion which utilized edge information to evaluate the significance of nodes. An innovative Iteration n-top strategy was then developed, guided by edge scores, to adaptively learn sparse hard clustering assignments for graphs. Additionally, a N-E Aggregation strategy is designed to aggregate node and edge features in each independent subgraph. Extensive experiments on the multi-site public datasets demonstrate the superiority and robustness of the proposed model. More notably, EHCPool has the potential to probe different types of dysfunctional brain networks from a data-driven perspective. Method code: https://github.com/swfen/EHCPool
Authors: Murat Onur Yildirim, Elif Ceren Gok Yildirim, Ghada Sokar, Decebal Constantin Mocanu, Joaquin Vanschoren
Continual learning (CL) refers to the ability of an intelligent system to sequentially acquire and retain knowledge from a stream of data with as little computational overhead as possible. To this end; regularization, replay, architecture, and parameter isolation approaches were introduced to the literature. Parameter isolation using a sparse network which enables to allocate distinct parts of the neural network to different tasks and also allows to share of parameters between tasks if they are similar. Dynamic Sparse Training (DST) is a prominent way to find these sparse networks and isolate them for each task. This paper is the first empirical study investigating the effect of different DST components under the CL paradigm to fill a critical research gap and shed light on the optimal configuration of DST for CL if it exists. Therefore, we perform a comprehensive study in which we investigate various DST components to find the best topology per task on well-known CIFAR100 and miniImageNet benchmarks in a task-incremental CL setup since our primary focus is to evaluate the performance of various DST criteria, rather than the process of mask selection. We found that, at a low sparsity level, Erdos-R\'enyi Kernel (ERK) initialization utilizes the backbone more efficiently and allows to effectively learn increments of tasks. At a high sparsity level, unless it is extreme, uniform initialization demonstrates a more reliable and robust performance. In terms of growth strategy; performance is dependent on the defined initialization strategy and the extent of sparsity. Finally, adaptivity within DST components is a promising way for better continual learners.
Authors: Lucas Brynte, José Pedro Iglesias, Carl Olsson, Fredrik Kahl
In this paper we tackle the problem of learning Structure-from-Motion (SfM) through the use of graph attention networks. SfM is a classic computer vision problem that is solved though iterative minimization of reprojection errors, referred to as Bundle Adjustment (BA), starting from a good initialization. In order to obtain a good enough initialization to BA, conventional methods rely on a sequence of sub-problems (such as pairwise pose estimation, pose averaging or triangulation) which provides an initial solution that can then be refined using BA. In this work we replace these sub-problems by learning a model that takes as input the 2D keypoints detected across multiple views, and outputs the corresponding camera poses and 3D keypoint coordinates. Our model takes advantage of graph neural networks to learn SfM-specific primitives, and we show that it can be used for fast inference of the reconstruction for new and unseen sequences. The experimental results show that the proposed model outperforms competing learning-based methods, and challenges COLMAP while having lower runtime.
Authors: Jiong Wang, Fengyu Yang, Wenbo Gou, Bingliang Li, Danqi Yan, Ailing Zeng, Yijun Gao, Junle Wang, Yanqing Jing, Ruimao Zhang
Estimating the 3D structure of the human body from natural scenes is a fundamental aspect of visual perception. 3D human pose estimation is a vital step in advancing fields like AIGC and human-robot interaction, serving as a crucial technique for understanding and interacting with human actions in real-world settings. However, the current datasets, often collected under single laboratory conditions using complex motion capture equipment and unvarying backgrounds, are insufficient. The absence of datasets on variable conditions is stalling the progress of this crucial task. To facilitate the development of 3D pose estimation, we present FreeMan, the first large-scale, multi-view dataset collected under the real-world conditions. FreeMan was captured by synchronizing 8 smartphones across diverse scenarios. It comprises 11M frames from 8000 sequences, viewed from different perspectives. These sequences cover 40 subjects across 10 different scenarios, each with varying lighting conditions. We have also established an semi-automated pipeline containing error detection to reduce the workload of manual check and ensure precise annotation. We provide comprehensive evaluation baselines for a range of tasks, underlining the significant challenges posed by FreeMan. Further evaluations of standard indoor/outdoor human sensing datasets reveal that FreeMan offers robust representation transferability in real and complex scenes. Code and data will be available at https://wangjiongw.github.io/freeman.
Authors: Manar Abdelatty, Joseph Incandela, Kangping Hu, Joseph W. Larkin, Sherief Reda, Jacob K. Rosenstein
Electrical capacitance tomography (ECT) is a nonoptical imaging technique in which a map of the interior permittivity of a volume is estimated by making capacitance measurements at its boundary and solving an inverse problem. While previous ECT demonstrations have often been at centimeter scales, ECT is not limited to macroscopic systems. In this paper, we demonstrate ECT imaging of polymer microspheres and bacterial biofilms using a CMOS microelectrode array, achieving spatial resolution of 10 microns. Additionally, we propose a deep learning architecture and an improved multi-objective training scheme for reconstructing out-of-plane permittivity maps from the sensor measurements. Experimental results show that the proposed approach is able to resolve microscopic 3-D structures, achieving 91.5% prediction accuracy on the microsphere dataset and 82.7% on the biofilm dataset, including an average of 4.6% improvement over baseline computational methods.
Authors: Abhishek Singh Sambyal, Usma Niyaz, Narayanan C. Krishnan, Deepti R. Bathula
In the field of medical image analysis, achieving high accuracy is not enough; ensuring well-calibrated predictions is also crucial. Confidence scores of a deep neural network play a pivotal role in explainability by providing insights into the model's certainty, identifying cases that require attention, and establishing trust in its predictions. Consequently, the significance of a well-calibrated model becomes paramount in the medical imaging domain, where accurate and reliable predictions are of utmost importance. While there has been a significant effort towards training modern deep neural networks to achieve high accuracy on medical imaging tasks, model calibration and factors that affect it remain under-explored. To address this, we conducted a comprehensive empirical study that explores model performance and calibration under different training regimes. We considered fully supervised training, which is the prevailing approach in the community, as well as rotation-based self-supervised method with and without transfer learning, across various datasets and architecture sizes. Multiple calibration metrics were employed to gain a holistic understanding of model calibration. Our study reveals that factors such as weight distributions and the similarity of learned representations correlate with the calibration trends observed in the models. Notably, models trained using rotation-based self-supervised pretrained regime exhibit significantly better calibration while achieving comparable or even superior performance compared to fully supervised models across different medical imaging datasets. These findings shed light on the importance of model calibration in medical image analysis and highlight the benefits of incorporating self-supervised learning approach to improve both performance and calibration.
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.
Authors: Tongtong Yuan, Xuange Zhang, Kun Liu, Bo Liu, Chen Chen, Jian Jin, Zhenzhen Jiao
Surveillance videos are an essential component of daily life with various critical applications, particularly in public security. However, current surveillance video tasks mainly focus on classifying and localizing anomalous events. Existing methods are limited to detecting and classifying the predefined events with unsatisfactory semantic understanding, although they have obtained considerable performance. To address this issue, we propose a new research direction of surveillance video-and-language understanding, and construct the first multimodal surveillance video dataset. We manually annotate the real-world surveillance dataset UCF-Crime with fine-grained event content and timing. Our newly annotated dataset, UCA (UCF-Crime Annotation), contains 23,542 sentences, with an average length of 20 words, and its annotated videos are as long as 110.7 hours. Furthermore, we benchmark SOTA models for four multimodal tasks on this newly created dataset, which serve as new baselines for surveillance video-and-language understanding. Through our experiments, we find that mainstream models used in previously publicly available datasets perform poorly on surveillance video, which demonstrates the new challenges in surveillance video-and-language understanding. To validate the effectiveness of our UCA, we conducted experiments on multimodal anomaly detection. The results demonstrate that our multimodal surveillance learning can improve the performance of conventional anomaly detection tasks. All the experiments highlight the necessity of constructing this dataset to advance surveillance AI. The link to our dataset is provided at: https://xuange923.github.io/Surveillance-Video-Understanding.
Authors: Kent Gauen, Stanley Chan
Efficiently computing attention maps for videos is challenging due to the motion of objects between frames. While a standard non-local search is high-quality for a window surrounding each query point, the window's small size cannot accommodate motion. Methods for long-range motion use an auxiliary network to predict the most similar key coordinates as offsets from each query location. However, accurately predicting this flow field of offsets remains challenging, even for large-scale networks. Small spatial inaccuracies significantly impact the attention module's quality. This paper proposes a search strategy that combines the quality of a non-local search with the range of predicted offsets. The method, named Shifted Non-Local Search, executes a small grid search surrounding the predicted offsets to correct small spatial errors. Our method's in-place computation consumes 10 times less memory and is over 3 times faster than previous work. Experimentally, correcting the small spatial errors improves the video frame alignment quality by over 3 dB PSNR. Our search upgrades existing space-time attention modules, which improves video denoising results by 0.30 dB PSNR for a 7.5% increase in overall runtime. We integrate our space-time attention module into a UNet-like architecture to achieve state-of-the-art results on video denoising.
Authors: Matt Allen, Francisco Dorr, Joseph A. Gallego-Mejia, Laura Martínez-Ferrer, Anna Jungbluth, Freddie Kalaitzis, Raúl Ramos-Pollán
In this work we pretrain a CLIP/ViT based model using three different modalities of satellite imagery across five AOIs covering over ~10\% of Earth's total landmass, namely Sentinel 2 RGB optical imagery, Sentinel 1 SAR radar amplitude and interferometric coherence. This model uses $\sim 250$ M parameters. Then, we use the embeddings produced for each modality with a classical machine learning method to attempt different downstream tasks for earth observation related to vegetation, built up surface, croplands and permanent water. We consistently show how we reduce the need for labeled data by 99\%, so that with ~200-500 randomly selected labeled examples (around 4K-10K km$^2$) we reach performance levels analogous to those achieved with the full labeled datasets (about 150K image chips or 3M km$^2$ in each area of interest - AOI) on all modalities, AOIs and downstream tasks. This leads us to think that the model has captured significant earth features useful in a wide variety of scenarios. To enhance our model's usability in practice, its architecture allows inference in contexts with missing modalities and even missing channels within each modality. Additionally, we visually show that this embedding space, obtained with no labels, is sensible to the different earth features represented by the labelled datasets we selected.
Authors: Matt Allen, Francisco Dorr, Joseph A. Gallego-Mejia, Laura Martínez-Ferrer, Anna Jungbluth, Freddie Kalaitzis, Raúl Ramos-Pollán
Satellite-based remote sensing is instrumental in the monitoring and mitigation of the effects of anthropogenic climate change. Large scale, high resolution data derived from these sensors can be used to inform intervention and policy decision making, but the timeliness and accuracy of these interventions is limited by use of optical data, which cannot operate at night and is affected by adverse weather conditions. Synthetic Aperture Radar (SAR) offers a robust alternative to optical data, but its associated complexities limit the scope of labelled data generation for traditional deep learning. In this work, we apply a self-supervised pretraining scheme, masked autoencoding, to SAR amplitude data covering 8.7\% of the Earth's land surface area, and tune the pretrained weights on two downstream tasks crucial to monitoring climate change - vegetation cover prediction and land cover classification. We show that the use of this pretraining scheme reduces labelling requirements for the downstream tasks by more than an order of magnitude, and that this pretraining generalises geographically, with the performance gain increasing when tuned downstream on regions outside the pretraining set. Our findings significantly advance climate change mitigation by facilitating the development of task and region-specific SAR models, allowing local communities and organizations to deploy tailored solutions for rapid, accurate monitoring of climate change effects.
Authors: Max Ku, Tianle Li, Kai Zhang, Yujie Lu, Xingyu Fu, Wenwen Zhuang, Wenhu Chen
Recently, a myriad of conditional image generation and editing models have been developed to serve different downstream tasks, including text-to-image generation, text-guided image editing, subject-driven image generation, control-guided image generation, etc. However, we observe huge inconsistencies in experimental conditions: datasets, inference, and evaluation metrics - render fair comparisons difficult. This paper proposes ImagenHub, which is a one-stop library to standardize the inference and evaluation of all the conditional image generation models. Firstly, we define seven prominent tasks and curate high-quality evaluation datasets for them. Secondly, we built a unified inference pipeline to ensure fair comparison. Thirdly, we design two human evaluation scores, i.e. Semantic Consistency and Perceptual Quality, along with comprehensive guidelines to evaluate generated images. We train expert raters to evaluate the model outputs based on the proposed metrics. Our human evaluation achieves a high inter-worker agreement of Krippendorff's alpha on 76% models with a value higher than 0.4. We comprehensively evaluated a total of around 30 models and observed three key takeaways: (1) the existing models' performance is generally unsatisfying except for Text-guided Image Generation and Subject-driven Image Generation, with 74% models achieving an overall score lower than 0.5. (2) we examined the claims from published papers and found 83% of them hold with a few exceptions. (3) None of the existing automatic metrics has a Spearman's correlation higher than 0.2 except subject-driven image generation. Moving forward, we will continue our efforts to evaluate newly published models and update our leaderboard to keep track of the progress in conditional image generation.
Authors: Bohan Zhai, Shijia Yang, Chenfeng Xu, Sheng Shen, Kurt Keutzer, Manling Li
Current large vision-language models (LVLMs) achieve remarkable progress, yet there remains significant uncertainty regarding their ability to accurately apprehend visual details, that is, in performing detailed captioning. To address this, we introduce $\textit{CCEval}$, a GPT-4 assisted evaluation method tailored for detailed captioning. Interestingly, while LVLMs demonstrate minimal object existence hallucination in existing VQA benchmarks, our proposed evaluation reveals continued susceptibility to such hallucinations. In this paper, we make the first attempt to investigate such hallucination from different aspects, including image resolution, the language decoder size, and instruction data amount, quality, granularity. Our findings underscore the unwarranted inference when the language description includes details at a finer object granularity than what the vision module can ground or verify, thus inducing hallucination. To control such hallucinations, we further attribute the reliability of captioning to contextual knowledge (involving only contextually grounded objects) and parametric knowledge (containing inferred objects by the model). Thus, we introduce $\textit{HallE-Switch}$, a controllable LVLM in terms of $\textbf{Hall}$ucination in object $\textbf{E}$xistence. HallE-Switch can condition the captioning to shift between (i) exclusively depicting contextual knowledge for grounded objects and (ii) blending it with parametric knowledge to imagine inferred objects. Our method reduces hallucination by 44% compared to LLaVA$_{7B}$ and maintains the same object coverage.
Authors: Laura Martínez-Ferrer, Anna Jungbluth, Joseph A. Gallego-Mejia, Matt Allen, Francisco Dorr, Freddie Kalaitzis, Raúl Ramos-Pollán
In this work we pre-train a DINO-ViT based model using two Synthetic Aperture Radar datasets (S1GRD or GSSIC) across three regions (China, Conus, Europe). We fine-tune the models on smaller labeled datasets to predict vegetation percentage, and empirically study the connection between the embedding space of the models and their ability to generalize across diverse geographic regions and to unseen data. For S1GRD, embedding spaces of different regions are clearly separated, while GSSIC's overlaps. Positional patterns remain during fine-tuning, and greater distances in embeddings often result in higher errors for unfamiliar regions. With this, our work increases our understanding of generalizability for self-supervised models applied to remote sensing.
Authors: Joseph A. Gallego-Mejia, Anna Jungbluth, Laura Martínez-Ferrer, Matt Allen, Francisco Dorr, Freddie Kalaitzis, Raúl Ramos-Pollán
Self-supervised learning (SSL) models have recently demonstrated remarkable performance across various tasks, including image segmentation. This study delves into the emergent characteristics of the Self-Distillation with No Labels (DINO) algorithm and its application to Synthetic Aperture Radar (SAR) imagery. We pre-train a vision transformer (ViT)-based DINO model using unlabeled SAR data, and later fine-tune the model to predict high-resolution land cover maps. We rigorously evaluate the utility of attention maps generated by the ViT backbone and compare them with the model's token embedding space. We observe a small improvement in model performance with pre-training compared to training from scratch and discuss the limitations and opportunities of SSL for remote sensing and land cover segmentation. Beyond small performance increases, we show that ViT attention maps hold great intrinsic value for remote sensing, and could provide useful inputs to other algorithms. With this, our work lays the groundwork for bigger and better SSL models for Earth Observation.
Authors: Yuxiu Lin, Hui Liu, Ren Wang, Qiang Guo, Caiming Zhang
Recent multi-view subspace clustering achieves impressive results utilizing deep networks, where the self-expressive correlation is typically modeled by a fully connected (FC) layer. However, they still suffer from two limitations. i) The parameter scale of the FC layer is quadratic to sample numbers, resulting in high time and memory costs that significantly degrade their feasibility in large-scale datasets. ii) It is under-explored to extract a unified representation that simultaneously satisfies minimal sufficiency and discriminability. To this end, we propose a novel deep framework, termed Efficient and Effective deep Multi-View Subspace Clustering (E$^2$MVSC). Instead of a parameterized FC layer, we design a Relation-Metric Net that decouples network parameter scale from sample numbers for greater computational efficiency. Most importantly, the proposed method devises a multi-type auto-encoder to explicitly decouple consistent, complementary, and superfluous information from every view, which is supervised by a soft clustering assignment similarity constraint. Following information bottleneck theory and the maximal coding rate reduction principle, a sufficient yet minimal unified representation can be obtained, as well as pursuing intra-cluster aggregation and inter-cluster separability within it. Extensive experiments show that E$^2$MVSC yields comparable results to existing methods and achieves state-of-the-art performance in various types of multi-view datasets.
Authors: Chaoyi Wu, Jiayu Lei, Qiaoyu Zheng, Weike Zhao, Weixiong Lin, Xiaoman Zhang, Xiao Zhou, Ziheng Zhao, Ya Zhang, Yanfeng Wang, Weidi Xie
Driven by the large foundation models, the development of artificial intelligence has witnessed tremendous progress lately, leading to a surge of general interest from the public. In this study, we aim to assess the performance of OpenAI's newest model, GPT-4V(ision), specifically in the realm of multimodal medical diagnosis. Our evaluation encompasses 17 human body systems, including Central Nervous System, Head and Neck, Cardiac, Chest, Hematology, Hepatobiliary, Gastrointestinal, Urogenital, Gynecology, Obstetrics, Breast, Musculoskeletal, Spine, Vascular, Oncology, Trauma, Pediatrics, with images taken from 8 modalities used in daily clinic routine, e.g., X-ray, Computed Tomography (CT), Magnetic Resonance Imaging (MRI), Positron Emission Tomography (PET), Digital Subtraction Angiography (DSA), Mammography, Ultrasound, and Pathology. We probe the GPT-4V's ability on multiple clinical tasks with or without patent history provided, including imaging modality and anatomy recognition, disease diagnosis, report generation, disease localisation.
Our observation shows that, while GPT-4V demonstrates proficiency in distinguishing between medical image modalities and anatomy, it faces significant challenges in disease diagnosis and generating comprehensive reports. These findings underscore that while large multimodal models have made significant advancements in computer vision and natural language processing, it remains far from being used to effectively support real-world medical applications and clinical decision-making.
All images used in this report can be found in https://github.com/chaoyi-wu/GPT-4V_Medical_Evaluation.
Authors: Pedram Agand, Mohammad Mahdavian, Manolis Savva, Mo Chen
In end-to-end autonomous driving, the utilization of existing sensor fusion techniques and navigational control methods for imitation learning proves inadequate in challenging situations that involve numerous dynamic agents. To address this issue, we introduce LeTFuser, a lightweight transformer-based algorithm for fusing multiple RGB-D camera representations. To perform perception and control tasks simultaneously, we utilize multi-task learning. Our model comprises of two modules, the first being the perception module that is responsible for encoding the observation data obtained from the RGB-D cameras. Our approach employs the Convolutional vision Transformer (CvT) \cite{wu2021cvt} to better extract and fuse features from multiple RGB cameras due to local and global feature extraction capability of convolution and transformer modules, respectively. Encoded features combined with static and dynamic environments are later employed by our control module to predict waypoints and vehicular controls (e.g. steering, throttle, and brake). We use two methods to generate the vehicular controls levels. The first method uses a PID algorithm to follow the waypoints on the fly, whereas the second one directly predicts the control policy using the measurement features and environmental state. We evaluate the model and conduct a comparative analysis with recent models on the CARLA simulator using various scenarios, ranging from normal to adversarial conditions, to simulate real-world scenarios. Our method demonstrated better or comparable results with respect to our baselines in term of driving abilities. The code is available at \url{https://github.com/pagand/e2etransfuser/tree/cvpr-w} to facilitate future studies.
Authors: Bo Ai, Zhanxin Wu, David Hsu
The data-driven approach to robot control has been gathering pace rapidly, yet generalization to unseen task domains remains a critical challenge. We argue that the key to generalization is representations that are (i) rich enough to capture all task-relevant information and (ii) invariant to superfluous variability between the training and the test domains. We experimentally study such a representation -- containing both depth and semantic information -- for visual navigation and show that it enables a control policy trained entirely in simulated indoor scenes to generalize to diverse real-world environments, both indoors and outdoors. Further, we show that our representation reduces the A-distance between the training and test domains, improving the generalization error bound as a result. Our proposed approach is scalable: the learned policy improves continuously, as the foundation models that it exploits absorb more diverse data during pre-training.
Authors: Fan Yang, Xiaofei Wang
Using deep learning methods to detect students' classroom behavior automatically is a promising approach for analyzing their class performance and improving teaching effectiveness. However, the lack of publicly available spatio-temporal datasets on student behavior, as well as the high cost of manually labeling such datasets, pose significant challenges for researchers in this field. To address this issue, we proposed a method for extending the spatio-temporal behavior dataset in Student Classroom Scenarios (SCB-ST-Dataset4) through image dataset. Our SCB-ST-Dataset4 comprises 757265 images with 25810 labels, focusing on 3 behaviors: hand-raising, reading, writing. Our proposed method can rapidly generate spatio-temporal behavior datasets without requiring extra manual labeling. Furthermore, we proposed a Behavior Similarity Index (BSI) to explore the similarity of behaviors. We evaluated the dataset using the YOLOv5, YOLOv7, YOLOv8, and SlowFast algorithms, achieving a mean average precision (map) of up to 82.3%. Last, we fused multiple models to generate student behavior-related data from various perspectives. The experiment further demonstrates the effectiveness of our method. And SCB-ST-Dataset4 provides a robust foundation for future research in student behavior detection, potentially contributing to advancements in this field. The SCB-ST-Dataset4 is available for download at: https://github.com/Whiffe/SCB-dataset.
Authors: Benjamin Salmon, Alexander Krull
Traditional supervised denoisers are trained using pairs of noisy input and clean target images. They learn to predict a central tendency of the posterior distribution over possible clean images. When, e.g., trained with the popular quadratic loss function, the network's output will correspond to the minimum mean square error (MMSE) estimate. Unsupervised denoisers based on Variational AutoEncoders (VAEs) have succeeded in achieving state-of-the-art results while requiring only unpaired noisy data as training input. In contrast to the traditional supervised approach, unsupervised denoisers do not directly produce a single prediction, such as the MMSE estimate, but allow us to draw samples from the posterior distribution of clean solutions corresponding to the noisy input. To approximate the MMSE estimate during inference, unsupervised methods have to create and draw a large number of samples - a computationally expensive process - rendering the approach inapplicable in many situations. Here, we present an alternative approach that trains a deterministic network alongside the VAE to directly predict a central tendency. Our method achieves results that surpass the results achieved by the unsupervised method at a fraction of the computational cost.
Authors: Qihang Zhou, Guansong Pang, Yu Tian, Shibo He, Jiming Chen
Zero-shot anomaly detection (ZSAD) requires detection models trained using auxiliary data to detect anomalies without any training sample in a target dataset. It is a crucial task when training data is not accessible due to various concerns, \eg, data privacy, yet it is challenging since the models need to generalize to anomalies across different domains where the appearance of foreground objects, abnormal regions, and background features, such as defects/tumors on different products/organs, can vary significantly. Recently large pre-trained vision-language models (VLMs), such as CLIP, have demonstrated strong zero-shot recognition ability in various vision tasks, including anomaly detection. However, their ZSAD performance is weak since the VLMs focus more on modeling the class semantics of the foreground objects rather than the abnormality/normality in the images. In this paper we introduce a novel approach, namely AnomalyCLIP, to adapt CLIP for accurate ZSAD across different domains. The key insight of AnomalyCLIP is to learn object-agnostic text prompts that capture generic normality and abnormality in an image regardless of its foreground objects. This allows our model to focus on the abnormal image regions rather than the object semantics, enabling generalized normality and abnormality recognition on diverse types of objects. Large-scale experiments on 17 real-world anomaly detection datasets show that AnomalyCLIP achieves superior zero-shot performance of detecting and segmenting anomalies in datasets of highly diverse class semantics from various defect inspection and medical imaging domains. Code will be made available at https://github.com/zqhang/AnomalyCLIP.
Authors: Mohammad Mehdi Rastikerdar, Jin Huang, Shiwei Fang, Hui Guan, Deepak Ganesan
While existing strategies to execute deep learning-based classification on low-power platforms assume the models are trained on all classes of interest, this paper posits that adopting context-awareness i.e. narrowing down a classification task to the current deployment context consisting of only recent inference queries can substantially enhance performance in resource-constrained environments. We propose a new paradigm, CACTUS, for scalable and efficient context-aware classification where a micro-classifier recognizes a small set of classes relevant to the current context and, when context change happens (e.g., a new class comes into the scene), rapidly switches to another suitable micro-classifier. CACTUS features several innovations, including optimizing the training cost of context-aware classifiers, enabling on-the-fly context-aware switching between classifiers, and balancing context switching costs and performance gains via simple yet effective switching policies. We show that CACTUS achieves significant benefits in accuracy, latency, and compute budget across a range of datasets and IoT platforms.
Authors: Zekun Wu, Shahin Doroudian, Aidong Lu
The understanding of visual analytics process can benefit visualization researchers from multiple aspects, including improving visual designs and developing advanced interaction functions. However, the log files of user behaviors are still hard to analyze due to the complexity of sensemaking and our lack of knowledge on the related user behaviors. This work presents a study on a comprehensive data collection of user behaviors, and our analysis approach with time-series classification methods. We have chosen a classical visualization application, Covid-19 data analysis, with common analysis tasks covering geo-spatial, time-series and multi-attributes. Our user study collects user behaviors on a diverse set of visualization tasks with two comparable systems, desktop and immersive visualizations. We summarize the classification results with three time-series machine learning algorithms at two scales, and explore the influences of behavior features. Our results reveal that user behaviors can be distinguished during the process of visual analytics and there is a potentially strong association between the physical behaviors of users and the visualization tasks they perform. We also demonstrate the usage of our models by interpreting open sessions of visual analytics, which provides an automatic way to study sensemaking without tedious manual annotations.
Authors: André Luiz Buarque Vieira e Silva, Heitor de Castro Felix, Franscisco Paulo Magalhães Simões, Veronica Teichrieb, Michel Mozinho dos Santos, Hemir Santiago, Virginia Sgotti, Henrique Lott Neto
Power line maintenance and inspection are essential to avoid power supply interruptions, reducing its high social and financial impacts yearly. Automating power line visual inspections remains a relevant open problem for the industry due to the lack of public real-world datasets of power line components and their various defects to foster new research. This paper introduces InsPLAD, a Power Line Asset Inspection Dataset and Benchmark containing 10,607 high-resolution Unmanned Aerial Vehicles colour images. The dataset contains seventeen unique power line assets captured from real-world operating power lines. Additionally, five of those assets present six defects: four of which are corrosion, one is a broken component, and one is a bird's nest presence. All assets were labelled according to their condition, whether normal or the defect name found on an image level. We thoroughly evaluate state-of-the-art and popular methods for three image-level computer vision tasks covered by InsPLAD: object detection, through the AP metric; defect classification, through Balanced Accuracy; and anomaly detection, through the AUROC metric. InsPLAD offers various vision challenges from uncontrolled environments, such as multi-scale objects, multi-size class instances, multiple objects per image, intra-class variation, cluttered background, distinct point-of-views, perspective distortion, occlusion, and varied lighting conditions. To the best of our knowledge, InsPLAD is the first large real-world dataset and benchmark for power line asset inspection with multiple components and defects for various computer vision tasks, with a potential impact to improve state-of-the-art methods in the field. It will be publicly available in its integrity on a repository with a thorough description. It can be found at https://github.com/andreluizbvs/InsPLAD.
Authors: Anna Reithmeir, Julia A. Schnabel, Veronika A. Zimmer
Medical image registration aims at identifying the spatial deformation between images of the same anatomical region and is fundamental to image-based diagnostics and therapy. To date, the majority of the deep learning-based registration methods employ regularizers that enforce global spatial smoothness, e.g., the diffusion regularizer. However, such regularizers are not tailored to the data and might not be capable of reflecting the complex underlying deformation. In contrast, physics-inspired regularizers promote physically plausible deformations. One such regularizer is the linear elastic regularizer which models the deformation of elastic material. These regularizers are driven by parameters that define the material's physical properties. For biological tissue, a wide range of estimations of such parameters can be found in the literature and it remains an open challenge to identify suitable parameter values for successful registration. To overcome this problem and to incorporate physical properties into learning-based registration, we propose to use a hypernetwork that learns the effect of the physical parameters of a physics-inspired regularizer on the resulting spatial deformation field. In particular, we adapt the HyperMorph framework to learn the effect of the two elasticity parameters of the linear elastic regularizer. Our approach enables the efficient discovery of suitable, data-specific physical parameters at test time.
Authors: Jun Myeong Choi, Max Christman, Roni Sengupta
In this paper, we develop a personalized video relighting algorithm that produces high-quality and temporally consistent relit video under any pose, expression, and lighting conditions in real-time. Existing relighting algorithms typically rely either on publicly available synthetic data, which yields poor relighting results, or instead on Light Stage data which is inaccessible and is not publicly available. We show that by casually capturing video of a user watching YouTube videos on a monitor we can train a personalized algorithm capable of producing high-quality relighting under any condition. Our key contribution is a novel neural relighting architecture that effectively separates the intrinsic appearance features, geometry and reflectance, from the source lighting and then combines it with the target lighting to generate a relit image. This neural architecture enables smoothing of intrinsic appearance features leading to temporally stable video relighting. Both qualitative and quantitative evaluations show that our relighting architecture improves portrait image relighting quality and temporal consistency over state-of-the-art approaches on both casually captured Light Stage at Your Desk (LSYD) data and Light Stage captured One Light At a Time (OLAT) datasets.
Authors: Yixun Liang, Xin Yang, Jiantao Lin, Haodong Li, Xiaogang Xu, Yingcong Chen
The recent advancements in text-to-3D generation mark a significant milestone in generative models, unlocking new possibilities for creating imaginative 3D assets across various real-world scenarios. While recent advancements in text-to-3D generation have shown promise, they often fall short in rendering detailed and high-quality 3D models. This problem is especially prevalent as many methods base themselves on Score Distillation Sampling (SDS). This paper identifies a notable deficiency in SDS, that it brings inconsistent and low-quality updating direction for the 3D model, causing the over-smoothing effect. To address this, we propose a novel approach called Interval Score Matching (ISM). ISM employs deterministic diffusing trajectories and utilizes interval-based score matching to counteract over-smoothing. Furthermore, we incorporate 3D Gaussian Splatting into our text-to-3D generation pipeline. Extensive experiments show that our model largely outperforms the state-of-the-art in quality and training efficiency.
Authors: Indu Ilanchezian, Valentyn Boreiko, Laura Kühlewein, Ziwei Huang, Murat Seçkin Ayhan, Matthias Hein, Lisa Koch, Philipp Berens
Counterfactual reasoning is often used in clinical settings to explain decisions or weigh alternatives. Therefore, for imaging based specialties such as ophthalmology, it would be beneficial to be able to create counterfactual images, illustrating answers to questions like "If the subject had had diabetic retinopathy, how would the fundus image have looked?". Here, we demonstrate that using a diffusion model in combination with an adversarially robust classifier trained on retinal disease classification tasks enables the generation of highly realistic counterfactuals of retinal fundus images and optical coherence tomography (OCT) B-scans. The key to the realism of counterfactuals is that these classifiers encode salient features indicative for each disease class and can steer the diffusion model to depict disease signs or remove disease-related lesions in a realistic way. In a user study, domain experts also found the counterfactuals generated using our method significantly more realistic than counterfactuals generated from a previous method, and even indistinguishable from real images.
Authors: Cristian Tommasino, Cristiano Russo, Antonio Maria Rinaldi, Francesco Ciompi
We present HoVer-UNet, an approach to distill the knowledge of the multi-branch HoVerNet framework for nuclei instance segmentation and classification in histopathology. We propose a compact, streamlined single UNet network with a Mix Vision Transformer backbone, and equip it with a custom loss function to optimally encode the distilled knowledge of HoVerNet, reducing computational requirements without compromising performances. We show that our model achieved results comparable to HoVerNet on the public PanNuke and Consep datasets with a three-fold reduction in inference time. We make the code of our model publicly available at https://github.com/DIAGNijmegen/HoVer-UNet.
Authors: Gauransh Sawhney, Daksh Dave, Adeel Ahmed, Jiechao Gao, Khalid Saleem
Unsupervised Domain Adaptation (UDA) methods facilitate knowledge transfer from a labeled source domain to an unlabeled target domain, navigating the obstacle of domain shift. While Convolutional Neural Networks (CNNs) are a staple in UDA, the rise of Vision Transformers (ViTs) provides new avenues for domain generalization. This paper presents an innovative method to bolster ViT performance in source-free target adaptation, beginning with an evaluation of how key, query, and value elements affect ViT outcomes. Experiments indicate that altering the key component has negligible effects on Transformer performance. Leveraging this discovery, we introduce Domain Representation Images (DRIs), feeding embeddings through the key element. DRIs act as domain-specific markers, effortlessly merging with the training regimen. To assess our method, we perform target adaptation tests on the Cross Instance DRI source-only (SO) control. We measure the efficacy of target adaptation with and without DRIs, against existing benchmarks like SHOT-B* and adaptations via CDTrans. Findings demonstrate that excluding DRIs offers limited gains over SHOT-B*, while their inclusion in the key segment boosts average precision promoting superior domain generalization. This research underscores the vital role of DRIs in enhancing ViT efficiency in UDA scenarios, setting a precedent for further domain adaptation explorations.
Authors: Antoine Guédon, Vincent Lepetit
We propose a method to allow precise and extremely fast mesh extraction from 3D Gaussian Splatting. Gaussian Splatting has recently become very popular as it yields realistic rendering while being significantly faster to train than NeRFs. It is however challenging to extract a mesh from the millions of tiny 3D gaussians as these gaussians tend to be unorganized after optimization and no method has been proposed so far. Our first key contribution is a regularization term that encourages the gaussians to align well with the surface of the scene. We then introduce a method that exploits this alignment to extract a mesh from the Gaussians using Poisson reconstruction, which is fast, scalable, and preserves details, in contrast to the Marching Cubes algorithm usually applied to extract meshes from Neural SDFs. Finally, we introduce an optional refinement strategy that binds gaussians to the surface of the mesh, and jointly optimizes these Gaussians and the mesh through Gaussian splatting rendering. This enables easy editing, sculpting, rigging, animating, compositing and relighting of the Gaussians using traditional softwares by manipulating the mesh instead of the gaussians themselves. Retrieving such an editable mesh for realistic rendering is done within minutes with our method, compared to hours with the state-of-the-art methods on neural SDFs, while providing a better rendering quality. Our project page is the following: https://anttwo.github.io/sugar/
Authors: Gengyuan Zhang, Jinhe Bi, Jindong Gu, Yanyu Chen, Volker Tresp
Understanding videos is an important research topic for multimodal learning. Leveraging large-scale datasets of web-crawled video-text pairs as weak supervision has become a pre-training paradigm for learning joint representations and showcased remarkable potential in video understanding tasks. However, videos can be multi-event and multi-grained, while these video-text pairs usually contain only broad-level video captions. This raises a question: with such weak supervision, can video representation in video-language models gain the ability to distinguish even factual discrepancies in textual description and understand fine-grained events? To address this, we introduce SPOT Prober, to benchmark existing video-language models's capacities of distinguishing event-level discrepancies as an indicator of models' event understanding ability. Our approach involves extracting events as tuples (<Subject, Predicate, Object, Attribute, Timestamps>) from videos and generating false event tuples by manipulating tuple components systematically. We reevaluate the existing video-language models with these positive and negative captions and find they fail to distinguish most of the manipulated events. Based on our findings, we propose to plug in these manipulated event captions as hard negative samples and find them effective in enhancing models for event understanding.
Authors: Jaeyoung Chung, Jeongtaek Oh, Kyoung Mu Lee
In this paper, we present a method to optimize Gaussian splatting with a limited number of images while avoiding overfitting. Representing a 3D scene by combining numerous Gaussian splats has yielded outstanding visual quality. However, it tends to overfit the training views when only a small number of images are available. To address this issue, we introduce a dense depth map as a geometry guide to mitigate overfitting. We obtained the depth map using a pre-trained monocular depth estimation model and aligning the scale and offset using sparse COLMAP feature points. The adjusted depth aids in the color-based optimization of 3D Gaussian splatting, mitigating floating artifacts, and ensuring adherence to geometric constraints. We verify the proposed method on the NeRF-LLFF dataset with varying numbers of few images. Our approach demonstrates robust geometry compared to the original method that relies solely on images. Project page: robot0321.github.io/DepthRegGS
Authors: Zhijie Rao, Jingcai Guo, Xiaocheng Lu, Qihua Zhou, Jie Zhang, Kang Wei, Chenxin Li, Song Guo
Generalized Zero-shot Learning (GZSL) has yielded remarkable performance by designing a series of unbiased visual-semantics mappings, wherein, the precision relies heavily on the completeness of extracted visual features from both seen and unseen classes. However, as a common practice in GZSL, the pre-trained feature extractor may easily exhibit difficulty in capturing domain-specific traits of the downstream tasks/datasets to provide fine-grained discriminative features, i.e., domain bias, which hinders the overall recognition performance, especially for unseen classes. Recent studies partially address this issue by fine-tuning feature extractors, while may inevitably incur catastrophic forgetting and overfitting issues. In this paper, we propose a simple yet effective Attribute-Aware Representation Rectification framework for GZSL, dubbed $\mathbf{(AR)^{2}}$, to adaptively rectify the feature extractor to learn novel features while keeping original valuable features. Specifically, our method consists of two key components, i.e., Unseen-Aware Distillation (UAD) and Attribute-Guided Learning (AGL). During training, UAD exploits the prior knowledge of attribute texts that are shared by both seen/unseen classes with attention mechanisms to detect and maintain unseen class-sensitive visual features in a targeted manner, and meanwhile, AGL aims to steer the model to focus on valuable features and suppress them to fit noisy elements in the seen classes by attribute-guided representation learning. Extensive experiments on various benchmark datasets demonstrate the effectiveness of our method.
Authors: Huanxin Chen, Pengshuai Yin, Huichou Huang, Qingyao Wu, Ruirui Liu, Xiatian Zhu
Predicting typhoon intensity accurately across space and time is crucial for issuing timely disaster warnings and facilitating emergency response. This has vast potential for minimizing life losses and property damages as well as reducing economic and environmental impacts. Leveraging satellite imagery for scenario analysis is effective but also introduces additional challenges due to the complex relations among clouds and the highly dynamic context. Existing deep learning methods in this domain rely on convolutional neural networks (CNNs), which suffer from limited per-layer receptive fields. This limitation hinders their ability to capture long-range dependencies and global contextual knowledge during inference. In response, we introduce a novel approach, namely "Typhoon Intensity Transformer" (Tint), which leverages self-attention mechanisms with global receptive fields per layer. Tint adopts a sequence-to-sequence feature representation learning perspective. It begins by cutting a given satellite image into a sequence of patches and recursively employs self-attention operations to extract both local and global contextual relations between all patch pairs simultaneously, thereby enhancing per-patch feature representation learning. Extensive experiments on a publicly available typhoon benchmark validate the efficacy of Tint in comparison with both state-of-the-art deep learning and conventional meteorological methods. Our code is available at https://github.com/chen-huanxin/Tint.
Authors: Shutong Zhang, Yi-Ling Qiao, Guanglei Zhu, Eric Heiden, Dylan Turpin, Jingzhou Liu, Ming Lin, Miles Macklin, Animesh Garg
Various heuristic objectives for modeling hand-object interaction have been proposed in past work. However, due to the lack of a cohesive framework, these objectives often possess a narrow scope of applicability and are limited by their efficiency or accuracy. In this paper, we propose HandyPriors, a unified and general pipeline for pose estimation in human-object interaction scenes by leveraging recent advances in differentiable physics and rendering. Our approach employs rendering priors to align with input images and segmentation masks along with physics priors to mitigate penetration and relative-sliding across frames. Furthermore, we present two alternatives for hand and object pose estimation. The optimization-based pose estimation achieves higher accuracy, while the filtering-based tracking, which utilizes the differentiable priors as dynamics and observation models, executes faster. We demonstrate that HandyPriors attains comparable or superior results in the pose estimation task, and that the differentiable physics module can predict contact information for pose refinement. We also show that our approach generalizes to perception tasks, including robotic hand manipulation and human-object pose estimation in the wild.
Authors: Peng Chen, Xiaobao Wei, Ming Lu, Yitong Zhu, Naiming Yao, Xingyu Xiao, Hui Chen
Speech-driven 3D facial animation has been an attractive task in both academia and industry. Traditional methods mostly focus on learning a deterministic mapping from speech to animation. Recent approaches start to consider the non-deterministic fact of speech-driven 3D face animation and employ the diffusion model for the task. However, personalizing facial animation and accelerating animation generation are still two major limitations of existing diffusion-based methods. To address the above limitations, we propose DiffusionTalker, a diffusion-based method that utilizes contrastive learning to personalize 3D facial animation and knowledge distillation to accelerate 3D animation generation. Specifically, to enable personalization, we introduce a learnable talking identity to aggregate knowledge in audio sequences. The proposed identity embeddings extract customized facial cues across different people in a contrastive learning manner. During inference, users can obtain personalized facial animation based on input audio, reflecting a specific talking style. With a trained diffusion model with hundreds of steps, we distill it into a lightweight model with 8 steps for acceleration. Extensive experiments are conducted to demonstrate that our method outperforms state-of-the-art methods. The code will be released.
Authors: Kunchang Li, Yali Wang, Yinan He, Yizhuo Li, Yi Wang, Yi Liu, Zun Wang, Jilan Xu, Guo Chen, Ping Luo, Limin Wang, Yu Qiao
With the rapid development of Multi-modal Large Language Models (MLLMs), a number of diagnostic benchmarks have recently emerged to evaluate the comprehension capabilities of these models. However, most benchmarks predominantly assess spatial understanding in the static image tasks, while overlooking temporal understanding in the dynamic video tasks. To alleviate this issue, we introduce a comprehensive Multi-modal Video understanding Benchmark, namely MVBench, which covers 20 challenging video tasks that cannot be effectively solved with a single frame. Specifically, we first introduce a novel static-to-dynamic method to define these temporal-related tasks. By transforming various static tasks into dynamic ones, we enable the systematic generation of video tasks that require a broad spectrum of temporal skills, ranging from perception to cognition. Then, guided by the task definition, we automatically convert public video annotations into multiple-choice QA to evaluate each task. On one hand, such a distinct paradigm allows us to build MVBench efficiently, without much manual intervention. On the other hand, it guarantees evaluation fairness with ground-truth video annotations, avoiding the biased scoring of LLMs. Moreover, we further develop a robust video MLLM baseline, i.e., VideoChat2, by progressive multi-modal training with diverse instruction-tuning data. The extensive results on our MVBench reveal that, the existing MLLMs are far from satisfactory in temporal understanding, while our VideoChat2 largely surpasses these leading models by over 15% on MVBench. All models and data are available at https://github.com/OpenGVLab/Ask-Anything.
Authors: Danah Yatim, Rafail Fridman, Omer Bar-Tal, Yoni Kasten, Tali Dekel
We present a new method for text-driven motion transfer - synthesizing a video that complies with an input text prompt describing the target objects and scene while maintaining an input video's motion and scene layout. Prior methods are confined to transferring motion across two subjects within the same or closely related object categories and are applicable for limited domains (e.g., humans). In this work, we consider a significantly more challenging setting in which the target and source objects differ drastically in shape and fine-grained motion characteristics (e.g., translating a jumping dog into a dolphin). To this end, we leverage a pre-trained and fixed text-to-video diffusion model, which provides us with generative and motion priors. The pillar of our method is a new space-time feature loss derived directly from the model. This loss guides the generation process to preserve the overall motion of the input video while complying with the target object in terms of shape and fine-grained motion traits.
Authors: Weilin Wan, Zhiyang Dou, Taku Komura, Wenping Wang, Dinesh Jayaraman, Lingjie Liu
Controllable human motion synthesis is essential for applications in AR/VR, gaming, movies, and embodied AI. Existing methods often focus solely on either language or full trajectory control, lacking precision in synthesizing motions aligned with user-specified trajectories, especially for multi-joint control. To address these issues, we present TLControl, a new method for realistic human motion synthesis, incorporating both low-level trajectory and high-level language semantics controls. Specifically, we first train a VQ-VAE to learn a compact latent motion space organized by body parts. We then propose a Masked Trajectories Transformer to make coarse initial predictions of full trajectories of joints based on the learned latent motion space, with user-specified partial trajectories and text descriptions as conditioning. Finally, we introduce an efficient test-time optimization to refine these coarse predictions for accurate trajectory control. Experiments demonstrate that TLControl outperforms the state-of-the-art in trajectory accuracy and time efficiency, making it practical for interactive and high-quality animation generation.
Authors: Zhiwen Fan, Kevin Wang, Kairun Wen, Zehao Zhu, Dejia Xu, Zhangyang Wang
Recent advancements in real-time neural rendering using point-based techniques have paved the way for the widespread adoption of 3D representations. However, foundational approaches like 3D Gaussian Splatting come with a substantial storage overhead caused by growing the SfM points to millions, often demanding gigabyte-level disk space for a single unbounded scene, posing significant scalability challenges and hindering the splatting efficiency.
To address this challenge, we introduce LightGaussian, a novel method designed to transform 3D Gaussians into a more efficient and compact format. Drawing inspiration from the concept of Network Pruning, LightGaussian identifies Gaussians that are insignificant in contributing to the scene reconstruction and adopts a pruning and recovery process, effectively reducing redundancy in Gaussian counts while preserving visual effects. Additionally, LightGaussian employs distillation and pseudo-view augmentation to distill spherical harmonics to a lower degree, allowing knowledge transfer to more compact representations while maintaining reflectance. Furthermore, we propose a hybrid scheme, VecTree Quantization, to quantize all attributes, resulting in lower bitwidth representations with minimal accuracy losses.
In summary, LightGaussian achieves an averaged compression rate over 15x while boosting the FPS from 139 to 215, enabling an efficient representation of complex scenes on Mip-NeRF 360, Tank and Temple datasets.
Project website: https://lightgaussian.github.io/
Authors: Wang Zhu, Ishika Singh, Yuan Huang, Robin Jia, Jesse Thomason
Data augmentation via back-translation is common when pretraining Vision-and-Language Navigation (VLN) models, even though the generated instructions are noisy. But: does that noise matter? We find that nonsensical or irrelevant language instructions during pretraining can have little effect on downstream performance for both HAMT and VLN-BERT on R2R, and is still better than only using clean, human data. To underscore these results, we concoct an efficient augmentation method, Unigram + Object, which generates nonsensical instructions that nonetheless improve downstream performance. Our findings suggest that what matters for VLN R2R pretraining is the quantity of visual trajectories, not the quality of instructions.
Authors: Duomin Wang, Bin Dai, Yu Deng, Baoyuan Wang
In this study, our goal is to create interactive avatar agents that can autonomously plan and animate nuanced facial movements realistically, from both visual and behavioral perspectives. Given high-level inputs about the environment and agent profile, our framework harnesses LLMs to produce a series of detailed text descriptions of the avatar agents' facial motions. These descriptions are then processed by our task-agnostic driving engine into motion token sequences, which are subsequently converted into continuous motion embeddings that are further consumed by our standalone neural-based renderer to generate the final photorealistic avatar animations. These streamlined processes allow our framework to adapt to a variety of non-verbal avatar interactions, both monadic and dyadic. Our extensive study, which includes experiments on both newly compiled and existing datasets featuring two types of agents -- one capable of monadic interaction with the environment, and the other designed for dyadic conversation -- validates the effectiveness and versatility of our approach. To our knowledge, we advanced a leap step by combining LLMs and neural rendering for generalized non-verbal prediction and photo-realistic rendering of avatar agents.
Authors: Jiaqi Zhao, Zeyu Ding, Yong Zhou, Hancheng Zhu, Wenliang Du, Rui Yao, Abdulmotaleb El Saddik
Object instances in remote sensing images often distribute with multi-orientations, varying scales, and dense distribution. These issues bring challenges to end-to-end oriented object detectors including multi-scale features alignment and a large number of queries. To address these limitations, we propose an end-to-end oriented detector equipped with an efficient decoder, which incorporates two technologies, Rotated RoI attention (RRoI attention) and Selective Distinct Queries (SDQ). Specifically, RRoI attention effectively focuses on oriented regions of interest through a cross-attention mechanism and aligns multi-scale features. SDQ collects queries from intermediate decoder layers and then filters similar queries to obtain distinct queries. The proposed SDQ can facilitate the optimization of one-to-one label assignment, without introducing redundant initial queries or extra auxiliary branches. Extensive experiments on five datasets demonstrate the effectiveness of our method. Notably, our method achieves state-of-the-art performance on DIOR-R (67.31% mAP), DOTA-v1.5 (67.43% mAP), and DOTA-v2.0 (53.28% mAP) with the ResNet50 backbone.
Authors: Baorui Ma, Haoge Deng, Junsheng Zhou, Yu-Shen Liu, Tiejun Huang, Xinlong Wang
Text-to-3D generation by distilling pretrained large-scale text-to-image diffusion models has shown great promise but still suffers from inconsistent 3D geometric structures (Janus problems) and severe artifacts. The aforementioned problems mainly stem from 2D diffusion models lacking 3D awareness during the lifting. In this work, we present GeoDream, a novel method that incorporates explicit generalized 3D priors with 2D diffusion priors to enhance the capability of obtaining unambiguous 3D consistent geometric structures without sacrificing diversity or fidelity. Specifically, we first utilize a multi-view diffusion model to generate posed images and then construct cost volume from the predicted image, which serves as native 3D geometric priors, ensuring spatial consistency in 3D space. Subsequently, we further propose to harness 3D geometric priors to unlock the great potential of 3D awareness in 2D diffusion priors via a disentangled design. Notably, disentangling 2D and 3D priors allows us to refine 3D geometric priors further. We justify that the refined 3D geometric priors aid in the 3D-aware capability of 2D diffusion priors, which in turn provides superior guidance for the refinement of 3D geometric priors. Our numerical and visual comparisons demonstrate that GeoDream generates more 3D consistent textured meshes with high-resolution realistic renderings (i.e., 1024 $\times$ 1024) and adheres more closely to semantic coherence.
Authors: Silvan Weder, Francis Engelmann, Johannes L. Schönberger, Akihito Seki, Marc Pollefeys, Martin R. Oswald
We propose an online 3D semantic segmentation method that incrementally reconstructs a 3D semantic map from a stream of RGB-D frames. Unlike offline methods, ours is directly applicable to scenarios with real-time constraints, such as robotics or mixed reality. To overcome the inherent challenges of online methods, we make two main contributions. First, to effectively extract information from the input RGB-D video stream, we jointly estimate geometry and semantic labels per frame in 3D. A key focus of our approach is to reason about semantic entities both in the 2D input and the local 3D domain to leverage differences in spatial context and network architectures. Our method predicts 2D features using an off-the-shelf segmentation network. The extracted 2D features are refined by a lightweight 3D network to enable reasoning about the local 3D structure. Second, to efficiently deal with an infinite stream of input RGB-D frames, a subsequent network serves as a temporal expert predicting the incremental scene updates by leveraging 2D, 3D, and past information in a learned manner. These updates are then integrated into a global scene representation. Using these main contributions, our method can enable scenarios with real-time constraints and can scale to arbitrary scene sizes by processing and updating the scene only in a local region defined by the new measurement. Our experiments demonstrate improved results compared to existing online methods that purely operate in local regions and show that complementary sources of information can boost the performance. We provide a thorough ablation study on the benefits of different architectural as well as algorithmic design decisions. Our method yields competitive results on the popular ScanNet benchmark and SceneNN dataset.
Authors: Ping Chen, Xingpeng Zhang, Chengtao Zhou, Dichao Fan, Peng Tu, Le Zhang, Yanlin Qian
Convolution neural network is successful in pervasive vision tasks, including label distribution learning, which usually takes the form of learning an injection from the non-linear visual features to the well-defined labels. However, how the discrepancy between features is mapped to the label discrepancy is ambient, and its correctness is not guaranteed. To address these problems, we study the mathematical connection between feature and its label, presenting a general and simple framework for label distribution learning. We propose a so-called Triangular Distribution Transform (TDT) to build an injective function between feature and label, guaranteeing that any symmetric feature discrepancy linearly reflects the difference between labels. The proposed TDT can be used as a plug-in in mainstream backbone networks to address different label distribution learning tasks. Experiments on Facial Age Recognition, Illumination Chromaticity Estimation, and Aesthetics assessment show that TDT achieves on-par or better results than the prior arts.
Authors: Samuel Stevens, Jiaman Wu, Matthew J Thompson, Elizabeth G Campolongo, Chan Hee Song, David Edward Carlyn, Li Dong, Wasila M Dahdul, Charles Stewart, Tanya Berger-Wolf, Wei-Lun Chao, Yu Su
Images of the natural world, collected by a variety of cameras, from drones to individual phones, are increasingly abundant sources of biological information. There is an explosion of computational methods and tools, particularly computer vision, for extracting biologically relevant information from images for science and conservation. Yet most of these are bespoke approaches designed for a specific task and are not easily adaptable or extendable to new questions, contexts, and datasets. A vision model for general organismal biology questions on images is of timely need. To approach this, we curate and release TreeOfLife-10M, the largest and most diverse ML-ready dataset of biology images. We then develop BioCLIP, a foundation model for the tree of life, leveraging the unique properties of biology captured by TreeOfLife-10M, namely the abundance and variety of images of plants, animals, and fungi, together with the availability of rich structured biological knowledge. We rigorously benchmark our approach on diverse fine-grained biology classification tasks, and find that BioCLIP consistently and substantially outperforms existing baselines (by 17% to 20% absolute). Intrinsic evaluation reveals that BioCLIP has learned a hierarchical representation conforming to the tree of life, shedding light on its strong generalizability. Our code, models and data will be made available at https://github.com/Imageomics/bioclip.
Authors: Tianwei Yin, Michaël Gharbi, Richard Zhang, Eli Shechtman, Fredo Durand, William T. Freeman, Taesung Park
Diffusion models generate high-quality images but require dozens of forward passes. We introduce Distribution Matching Distillation (DMD), a procedure to transform a diffusion model into a one-step image generator with minimal impact on image quality. We enforce the one-step image generator match the diffusion model at distribution level, by minimizing an approximate KL divergence whose gradient can be expressed as the difference between 2 score functions, one of the target distribution and the other of the synthetic distribution being produced by our one-step generator. The score functions are parameterized as two diffusion models trained separately on each distribution. Combined with a simple regression loss matching the large-scale structure of the multi-step diffusion outputs, our method outperforms all published few-step diffusion approaches, reaching 2.62 FID on ImageNet 64x64 and 11.49 FID on zero-shot COCO-30k, comparable to Stable Diffusion but orders of magnitude faster. Utilizing FP16 inference, our model generates images at 20 FPS on modern hardware.
Authors: Zidu Wang, Xiangyu Zhu, Tianshuo Zhang, Baiqin Wang, Zhen Lei
3D Morphable Models (3DMMs) provide promising 3D face reconstructions in various applications. However, existing methods struggle to reconstruct faces with extreme expressions due to deficiencies in supervisory signals, such as sparse or inaccurate landmarks. Segmentation information contains effective geometric contexts for face reconstruction. Certain attempts intuitively depend on differentiable renderers to compare the rendered silhouettes of reconstruction with segmentation, which is prone to issues like local optima and gradient instability. In this paper, we fully utilize the facial part segmentation geometry by introducing Part Re-projection Distance Loss (PRDL). Specifically, PRDL transforms facial part segmentation into 2D points and re-projects the reconstruction onto the image plane. Subsequently, by introducing grid anchors and computing different statistical distances from these anchors to the point sets, PRDL establishes geometry descriptors to optimize the distribution of the point sets for face reconstruction. PRDL exhibits a clear gradient compared to the renderer-based methods and presents state-of-the-art reconstruction performance in extensive quantitative and qualitative experiments. The project will be publicly available.
Authors: Shaohua Dong, Yunhe Feng, Qing Yang, Yan Huang, Dongfang Liu, Heng Fan
Multimodal (e.g., RGB-Depth/RGB-Thermal) fusion has shown great potential for improving semantic segmentation in complex scenes (e.g., indoor/low-light conditions). Existing approaches often fully fine-tune a dual-branch encoder-decoder framework with a complicated feature fusion strategy for achieving multimodal semantic segmentation, which is training-costly due to the massive parameter updates in feature extraction and fusion. To address this issue, we propose a surprisingly simple yet effective dual-prompt learning network (dubbed DPLNet) for training-efficient multimodal (e.g., RGB-D/T) semantic segmentation. The core of DPLNet is to directly adapt a frozen pre-trained RGB model to multimodal semantic segmentation, reducing parameter updates. For this purpose, we present two prompt learning modules, comprising multimodal prompt generator (MPG) and multimodal feature adapter (MFA). MPG works to fuse the features from different modalities in a compact manner and is inserted from shadow to deep stages to generate the multi-level multimodal prompts that are injected into the frozen backbone, while MPG adapts prompted multimodal features in the frozen backbone for better multimodal semantic segmentation. Since both the MPG and MFA are lightweight, only a few trainable parameters (3.88M, 4.4% of the pre-trained backbone parameters) are introduced for multimodal feature fusion and learning. Using a simple decoder (3.27M parameters), DPLNet achieves new state-of-the-art performance or is on a par with other complex approaches on four RGB-D/T semantic segmentation datasets while satisfying parameter efficiency. Moreover, we show that DPLNet is general and applicable to other multimodal tasks such as salient object detection and video semantic segmentation. Without special design, DPLNet outperforms many complicated models. Our code will be available at github.com/ShaohuaDong2021/DPLNet.
Authors: Sangamesh Kodge, Gobinda Saha, Kaushik Roy
Machine unlearning has emerged as a prominent and challenging area of interest, driven in large part by the rising regulatory demands for industries to delete user data upon request and the heightened awareness of privacy. Existing approaches either retrain models from scratch or use several finetuning steps for every deletion request, often constrained by computational resource limitations and restricted access to the original training data. In this work, we introduce a novel class unlearning algorithm designed to strategically eliminate an entire class or a group of classes from the learned model. To that end, our algorithm first estimates the Retain Space and the Forget Space, representing the feature or activation spaces for samples from classes to be retained and unlearned, respectively. To obtain these spaces, we propose a novel singular value decomposition-based technique that requires layer wise collection of network activations from a few forward passes through the network. We then compute the shared information between these spaces and remove it from the forget space to isolate class-discriminatory feature space for unlearning. Finally, we project the model weights in the orthogonal direction of the class-discriminatory space to obtain the unlearned model. We demonstrate our algorithm's efficacy on ImageNet using a Vision Transformer with only $\sim$1.5% drop in retain accuracy compared to the original model while maintaining under 1% accuracy on the unlearned class samples. Further, our algorithm consistently performs well when subject to Membership Inference Attacks showing 7.8% improvement on average across a variety of image classification datasets and network architectures, as compared to other baselines while being $\sim$6x more computationally efficient.