Abstract State-space models (SSMs), exemplified by S4, have introduced a novel context modeling method by integrating state-space techniques into deep learning. Despite their effectiveness, SSMs struggle with global context modeling due to data-independent matrices. The Mamba model addresses this with data-dependent variants enabled by the S6 selective-scan algorithm, enhancing context modeling, especially for long sequences. However, Mamba-based architectures face significant parameter scalability challenges, limiting their utility in vision applications. This paper tackles the scalability issue of large SSMs for image classification and action recognition without relying on additional techniques like knowledge distillation. We analyze the distinct characteristics of Mamba-based and Attention-based models, proposing a Mamba-Attention interleaved architecture that enhances scalability, robustness, and performance. We demonstrate that the stable and efficient interleaved architecture resolves the scalability issue of Mamba-based architectures and increases robustness to common corruption artifacts. Our thorough evaluation on the ImageNet-1K, Kinetics-400, and Something-Something-v2 benchmarks demonstrates that our approach improves the accuracy of state-of-the-art Mamba-based architectures by up to $$+1.7$$ + 1.7 %.

Uncertainty-aware consistency learning is one of the reliable approaches in semi-supervised medical image segmentation, enforcing robust model predictions under various perturbations. However, existing methods often rely on multiple stochastic predictions or dual-network/decoder discrepancies to estimate uncertainty, which increases computational cost and discards uncertain regions, potentially missing complex structures such as ambiguous lesion boundaries. To address these challenges, we introduce a Dual Uncertainty-Guided Consistency and Regional Contrastive Learning (DUCore) framework. DUCore improves segmentation robustness by integrating two complementary loss functions within consistency learning. The dual uncertainty-guided consistency loss (DuCL) adaptively calibrates the prediction alignment by prioritizing uncertain regions. DuCL uses deterministic single-pass uncertainty estimation, employing entropy-based calibration for aleatoric uncertainty and Proxy Dirichlet calibration for epistemic uncertainty. These uncertainty measures are computed directly from network output, and moderately uncertain regions are weighted instead of being discarded, which preserves valuable learning signals. The Regional Contrastive Loss (ReCL) further refines feature separability using boundary- and gradient-based hard negative mining in the encoded representation space. By explicitly targeting structural ambiguities, ReCL distinguishes lesion and organ edges from visually similar boundary-adjacent regions and mitigates intensity overlaps in gradient-rich transitions. As a result, DUCore is able to delineate fine structures and complex boundaries with higher precision. Extensive experiments on various medical segmentation benchmarks reveal that DUCore outperforms existing consistency methods.

Foundational vision-language models like CLIP are emerging as a promising paradigm in vision due to their excellent generalization. However, adapting these models for downstream tasks while maintaining their generalization remains challenging. In literature, one branch of methods adapts CLIP by learning prompts using images. While effective, these methods often rely on image-label data, which is not always practical, and struggle to generalize to new datasets due to overfitting on few-shot source data. Another approach explores training-free methods by generating class captions from large language models (LLMs) and performing prompt ensembling, but these methods often produce static, class-specific prompts that cannot be transferred to new classes and incur additional costs by generating LLM descriptions for each class separately. In this work, we aim to combine the strengths of both approaches by learning prompts using only text data derived from LLMs. As supervised training of prompts in the image-free setup is non-trivial, we develop a language-only efficient training approach that enables prompts to distill rich contextual knowledge from LLM data. Furthermore, by mapping the LLM contextual text data within the learned prompts, our approach enables zero-shot transfer of prompts to new classes and datasets, potentially reducing the LLM prompt engineering cost. To the best of our knowledge, this is the first work that learns generalized and transferable prompts for image tasks using only text data. We perform evaluations on 4 benchmarks, where ProText improves over ensembling methods while being competitive with those using labeled images.

Capitalizing on a vast amount of image-text data, large-scale vision-language pre-training has demonstrated remarkable zero-shot capabilities and has been utilized in several applications. However, models trained on general everyday web-crawled data often exhibit sub-optimal performance for specialized domains, likely due to domain shift. Recent works have tackled this problem for some domains (e.g., healthcare) by constructing domain-specialized image-text data. However, constructing a dedicated large-scale image-text dataset for sustainable areas of agriculture and livestock is still open to research. Further, this domain desires fine-grained feature learning due to the subtle nature of the downstream tasks (e.g., nutrient deficiency detection and livestock breed classification). To address this, we present AgriCLIP, a vision-language foundational model dedicated to the domain of agriculture and livestock. First, we propose a large-scale dataset named ALive that leverages a customized prompt generation strategy to overcome the scarcity of expert annotations. Our ALive dataset covers crops, livestock, and fishery, with around 600,000 image-text pairs. Second, we propose a training pipeline that integrates both contrastive and self-supervised learning to learn both global semantic and local fine-grained domain-specialized features. Experiments on a diverse set of 20 downstream tasks demonstrate the effectiveness of the AgriCLIP framework, achieving an absolute gain of 9.07% in terms of average zero-shot classification accuracy over the standard CLIP adaptation via domain-specialized ALive dataset. Our ALive dataset and code can be accessible on Github.

Semi-supervised learning in medical image segmentation leverages unlabeled data to reduce annotation burdens through consistency learning. However, current methods struggle with class imbalance and high uncertainty from pathology variations, leading to inaccurate segmentation in 3D medical images. To address these challenges, we present DyCON, a Dynamic Uncertainty-aware Consistency and Contrastive Learning framework that enhances the generalization of consistency methods with two complementary losses: Uncertainty-aware Consistency Loss (UnCL) and Focal Entropy-aware Contrastive Loss (FeCL). UnCL enforces global consistency by dynamically weighting the contribution of each voxel to the consistency loss based on its uncertainty, preserving high-uncertainty regions instead of filtering them out. Initially, UnCL prioritizes learning from uncertain voxels with lower penalties, encouraging the model to explore challenging regions. As training progress, the penalty shift towards confident voxels to refine predictions and ensure global consistency. Meanwhile, FeCL enhances local feature discrimination in imbalanced regions by introducing dual focal mechanisms and adaptive confidence adjustments into the contrastive principle. These mechanisms jointly prioritizes hard positives and negatives while focusing on uncertain sample pairs, effectively capturing subtle lesion variations under class imbalance. Extensive evaluations on four diverse medical image segmentation datasets (ISLES’22, BraTS’19, LA, Pancreas) show DyCON’s superior performance against SOTA methods1.

The rapid proliferation of large-scale text-to-image diffusion (T2ID) models has raised serious concerns about their potential misuse in generating harmful content. Although numerous methods have been proposed for erasing undesired concepts from T2ID models, they often provide a false sense of security; concept-erased models (CEMs) can still be manipulated via adversarial attacks to regenerate the erased concept. While a few robust concept erasure methods based on adversarial training have emerged recently, they compromise on utility (generation quality for benign concepts) to achieve robustness and/or remain vulnerable to advanced embedding space attacks. These limitations stem from the failure of robust CEMs to thoroughly search for "blind spots" in the embedding space. To bridge this gap, we propose STEREO, a novel two-stage framework that employs adversarial training as a first step rather than the only step for robust concept erasure. In the first stage, STEREO employs adversarial training as a vulnerability identification mechanism to search thoroughly enough. In the second robustly erase once stage, STEREO introduces an anchor-concept-based compositional objective to robustly erase the target concept in a single fine-tuning stage, while minimizing the degradation of model utility. We benchmark STEREO against seven state-of-the-art concept erasure methods, demonstrating its superior robustness to both white-box and black-box attacks, while largely preserving utility.

Advancements in Computer-Aided Screening (CAS) systems are essential for improving the detection of security threats in X-ray baggage scans. However, current datasets are limited in representing real-world, sophisticated threats and concealment tactics, and existing approaches are constrained by a closed-set paradigm with predefined labels. To address these challenges, we introduce STCray, the first multimodal X-ray baggage security dataset, comprising 46,642 image-caption paired scans across 21 threat categories, generated using an X-ray scanner for airport security. STCray is meticulously developed with our specialized protocol that ensures domain-aware, coherent captions, that lead to the multi-modal instruction following data in X-ray baggage security. This allows us to train a domain-aware visual AI assistant named STING-BEE that supports a range of vision-language tasks, including scene comprehension, referring threat localization, visual grounding, and visual question answering (VQA), establishing novel baselines for multi-modal learning in X-ray baggage security. Further, STING-BEE shows state-of-the-art generalization in cross-domain settings. Code, data, and models are available at https://divs1159.github.io/STING-BEE/.

We present an efficient encoder-free approach for video-language understanding that achieves competitive performance while significantly reducing computational overhead. Current video-language models typically rely on heavyweight image encoders (300M-1.1B parameters) or video encoders (1B-1.4B parameters), creating a substantial computational burden when processing multi-frame videos. Our method introduces a novel Spatio-Temporal Alignment Block (STAB) that directly processes video inputs without requiring pre-trained encoders while using only 45M parameters for visual processing - at least a 6.5× reduction compared to traditional approaches. The STAB architecture combines Local Spatio-Temporal Encoding for fine-grained feature extraction, efficient spatial downsampling through learned attention and separate mechanisms for modeling frame-level and video-level relationships. Our model achieves comparable or superior performance to encoder-based approaches for open-ended video question answering on standard benchmarks. The fine-grained video question-answering evaluation demonstrates our model’s effectiveness, outperforming the encoder-based approaches Video-ChatGPT and Video-LLaVA in key aspects like correctness and temporal understanding. Extensive ablation studies validate our architectural choices and demonstrate the effectiveness of our spatio-temporal modeling approach while achieving 3-4× faster processing speeds than previous methods. Code is available at https://jh-yi.github.io/Video-Panda.

We present MixANT, a novel architecture for stochastic long-term dense anticipation of human activities. While recent State Space Models (SSMs) like Mamba have shown promise through input-dependent selectivity on three key parameters, the critical forget-gate ($\textbf{A}$ matrix) controlling temporal memory remains static. We address this limitation by introducing a mixture of experts approach that dynamically selects contextually relevant $\textbf{A}$ matrices based on input features, enhancing representational capacity without sacrificing computational efficiency. Extensive experiments on the 50Salads, Breakfast, and Assembly101 datasets demonstrate that MixANT consistently outperforms state-of-the-art methods across all evaluation settings. Our results highlight the importance of input-dependent forget-gate mechanisms for reliable prediction of human behavior in diverse real-world scenarios.

In this work, we address the challenging and emergent problem of novel object detection (NOD), focusing on the accurate detection of both known and novel object categories during inference. Traditional object detection algorithms are inherently closed-set, limiting their capability to handle NOD. We present a novel approach to transform existing closed-set detectors into open-set detectors. This transformation is achieved by leveraging the complementary strengths of pre-trained foundational models, specifically CLIP and SAM, through our cooperative mechanism. Furthermore, by integrating this mechanism with state-of-the-art open-set detectors such as GDINO, we establish new benchmarks in object detection performance. Our method achieves 17.42 mAP in novel object detection and 42.08 mAP for known objects on the challenging LVIS dataset. Adapting our approach to the COCO OVD split, we obtain an impressive result of 49.6 Novel AP50, which outperforms existing SOTA methods with similar backbone. Our code is available at: https://rohit901.github.io/coop-foundation-models/.

Vision systems that see and reason about the compositional nature of visual scenes are fundamental to understanding our world. The complex relations between objects and their locations, ambiguities, and variations in the real-world environment can be better described in human language, naturally governed by grammatical rules and other modalities such as audio and depth. The models learned to bridge the gap between such modalities and large-scale training data facilitate contextual reasoning, generalization, and prompt capabilities at test time. These models are referred to as foundation models. The output of such models can be modified through human-provided prompts without retraining, e.g., segmenting a particular object by providing a bounding box, having interactive dialogues by asking questions about an image or video scene or manipulating the robot's behavior through language instructions. In this survey, we provide a comprehensive review of such emerging foundation models, including typical architecture designs to combine different modalities (vision, text, audio, etc.), training objectives (contrastive, generative), pre-training datasets, fine-tuning mechanisms, and the common prompting patterns; textual, visual, and heterogeneous. We discuss the open challenges and research directions for foundation models in computer vision, including difficulties in their evaluations and benchmarking, gaps in their real-world understanding, limitations of contextual understanding, biases, vulnerability to adversarial attacks, and interpretability issues. We review recent developments in this field, covering a wide range of applications of foundation models systematically and comprehensively.

Most existing multi-object tracking methods typically learn visual tracking features via maximizing dis-similarities of different instances and minimizing similarities of the same instance. While such a feature learning scheme achieves promising performance, learning discriminative features solely based on visual information is challenging especially in case of environmental interference such as occlusion, blur and domain variance. In this work, we argue that multi-modal language-driven features provide complementary information to classical visual features, thereby aiding in improving the robustness to such environmental interference. To this end, we propose a new multi-object tracking framework, named LG-MOT, that explicitly leverages language information at different levels of granularity (scene-and instance-level) and combines it with standard visual features to obtain discriminative representations. To develop LG-MOT, we annotate existing MOT datasets with scene-and instance-level language descriptions. We then encode both scene-and instance-level language information into high-dimensional embeddings, which are utilized to guide the visual features during training. At inference, our LG-MOT uses the standard visual features without relying on annotated language descriptions. Extensive experiments on three benchmarks, MOT17, DanceTrack and SportsMOT, reveal the merits of the proposed contributions leading to state-of-the-art performance. On the DanceTrack test set, our LG-MOT achieves an absolute gain of 2.2% in terms of target object association (IDF1 score), compared to the baseline using only visual features. Further, our LG-MOT exhibits strong cross-domain generalizability. Source code and pre-trained models are available at https://github.com/WesLee88524/LG-MOT.

Large-scale pre-trained Vision Language Models (VLMs) have proven effective for zero-shot classification. Despite the success, most traditional VLMs-based methods are restricted by the assumption of partial source supervision or ideal target vocabularies, which rarely satisfy the open-world scenario. In this paper, we aim at a more challenging setting, Realistic Zero-Shot Classification, which assumes no annotation but instead a broad vocabulary. To address the new problem, we propose the Self Structural Semantic Alignment (S3A) framework, which extracts the structural semantic information from unlabeled data while simultaneously self-learning. Our S3A framework adopts a unique Cluster-Vote-Prompt-Realign (CVPR) algorithm, which iteratively groups unlabeled data to derive structural semantics for pseudo-supervision. Our CVPR algorithm includes iterative clustering on images, voting within each cluster to identify initial class candidates from the vocabulary, generating discriminative prompts with large language models to discern confusing candidates, and realigning images and the vocabulary as structural semantic alignment. Finally, we propose to self-train the CLIP image encoder with both individual and structural semantic alignment through a teacher-student learning strategy. Our comprehensive experiments across various generic and fine-grained benchmarks demonstrate that the S3A method substantially improves over existing VLMs-based approaches, achieving a more than 15% accuracy improvement over CLIP on average. Our codes, models, and prompts are publicly released at https://github.com/sheng-eatamath/S3A.

Recent advancements in Large Vision-Language Models (VLMs) have shown great promise in natural image domains, allowing users to hold a dialogue about given visual content. However, such general-domain VLMs perform poorly for Remote Sensing (RS) scenarios, leading to inaccurate or fabricated information when presented with RS domain-specific queries. Such a behavior emerges due to the unique challenges introduced by RS imagery. For example, to handle high-resolution RS imagery with diverse scale changes across categories and many small objects, region-level reasoning is necessary alongside holistic scene inter-pretation. Furthermore, the lack of domain-specific multimodal instruction following data as well as strong back-bone models for RS make it hard for the models to align their behavior with user queries. To address these limitations, we propose GeoChat - the first versatile remote sensing VLM that offers multitask conversational capabilities with high-resolution RS images. Specifically, GeoChat can not only answer image-level queries but also accepts region inputs to hold region-specific dialogue. Further-more, it can visually ground objects in its responses by referring to their spatial coordinates. To address the lack of domain-specific datasets, we generate a novel RS multimodal instruction-following dataset by extending image-text pairs from existing diverse RS datasets. We establish a comprehensive benchmarkfor RS multitask conversations and compare with a number of baseline methods. GeoChat demonstrates robust zero-shot performance on various RS tasks, e.g., image and region captioning, visual question answering, scene classification, visually grounded conversations and referring detection. Our code is available here.

Recent advances in unsupervised learning have demonstrated the ability of large vision models to achieve promising results on downstream tasks by pre-training on large amount of unlabelled data. Such pre-training techniques have also been explored recently in the remote sensing domain due to the availability of large amount of unlabelled data. Different from standard natural image datasets, remote sensing data is acquired from various sensor technologies and exhibit diverse range of scale variations as well as modalities. Existing satellite image pre-training methods either ignore the scale information present in the remote sensing imagery or restrict themselves to use only a single type of data modality. In this paper, we re-visit transformers pre-training and leverage multi-scale information that is effectively utilized with multiple modalities. Our proposed approach, named SatMAE++, performs multi-scale pre-training and utilizes convolution based upsampling blocks to reconstruct the image at higher scales making it extensible to include more scales. Compared to existing works, the proposed SatMAE++ with multi-scale pre-training is equally effective for both optical as well as multi-spectral imagery. Extensive experiments on six datasets reveal the merits of proposed contributions, leading to state-of-the-art performance on all datasets. Sat-MAE++ achieves mean average precision (mAP) gain of 2.5% for multi-label classification task on BigEarthNet dataset. Our code and pre-trained models are available at https://github.com/techmn/satmae_pp.

Composed video retrieval (CoVR) is a challenging problem in computer vision which has recently highlighted the integration of modification text with visual queries for more sophisticated video search in large databases. Existing works predominantly rely on visual queries combined with modification text to distinguish relevant videos. However, such a strategy struggles to fully preserve the rich query-specific context in retrieved target videos and only represents the target video using visual embedding. We introduce a novel CoVR framework that leverages detailed language descriptions to explicitly encode query-specific contextual information and learns discriminative embeddings of vision only, text only and vision-text for better alignment to accurately retrieve matched target videos. Our proposed framework can be flexibly employed for both composed video (CoVR) and image (CoIR) retrieval tasks. Experiments on three datasets show that our approach obtains state-of-the-art performance for both CovR and zero-shot CoIR tasks, achieving gains as high as around 7% in terms of recall@ K=1 score. Our code, detailed language descriptions for Web ViD-Co VR dataset are available at https://github.com/OmkarThawakar/composed-video-retrieval.

Video grounding aims to localize a spatio-temporal section in a video corresponding to an input text query. This paper addresses a critical limitation in current video grounding methodologies by introducing an Open-Vocabulary Spatio- Temporal Video Grounding task. Unlike prevalent closed-set approaches that struggle with open-vocabulary scenarios due to limited training data and pre-defined vocabularies, our model leverages pre-trained rep-resentations from foundational spatial grounding models. This empowers it to effectively bridge the semantic gap be-tween natural language and diverse visual content, achieving strong performance in closed-set and open-vocabulary settings. Our contributions include a novel spatio-temporal video grounding model, surpassing state-of-the-art results in closed-set evaluations on multiple datasets and demon-strating superior performance in open-vocabulary scenar-ios. Notably, the proposed model outperforms state-of-the-art methods in closed-set settings on VidSTG (Declarative and Interrogative) and HC-STVG (VI and V2) datasets. Furthermore, in open-vocabulary evaluations on HC-STVG VI and YouCook-Interactions, our model surpasses the re-cent best-performing models by 4.88 m.vloU and 1.83% ac-curacy, demonstrating its efficacy in handling diverse lin-guistic and visual concepts for improved video understanding. Our codes will be publicly released.

Diffusion-based generative models have significantly advanced text-to-image generation but encounter challenges when processing lengthy and intricate text prompts describing complex scenes with multiple objects. While excelling in generating images from short, single-object descriptions, these models often struggle to faithfully capture all the nuanced details within longer and more elaborate textual inputs. In response, we present a novel approach leveraging Large Language Models (LLMs) to extract critical components from text prompts, including bounding box coordinates for foreground objects, detailed textual descriptions for individual objects, and a succinct background context. These components form the foundation of our layout-to-image generation model, which operates in two phases. The initial Global Scene Generation utilizes object layouts and background context to create an initial scene but often falls short in faithfully representing object characteristics as specified in the prompts. To address this limitation, we introduce an Iterative Refinement Scheme that iteratively evaluates and refines box-level content to align them with their textual descriptions, recomposing objects as needed to ensure consistency. Our evaluation on complex prompts featuring multiple objects demonstrates a substantial improvement in recall compared to baseline diffusion models. This is further validated by a user study, underscoring the efficacy of our approach in generating coherent and detailed scenes from intricate textual inputs. Our iterative framework offers a promising solution for enhancing text-to-image generation models' fidelity with lengthy, multifaceted descriptions, opening new possibilities for accurate and diverse image synthesis from textual inputs.

Adversarial training (AT) is an effective approach to making deep neural networks robust against adversarial attacks. Recently, different AT defenses are proposed that not only maintain a high clean accuracy but also show significant robustness against popular and well-studied adversarial attacks, such as projected gradient descent (PGD). High adversarial robustness can also arise if an attack fails to find adversarial gradient directions, a phenomenon known as "gradient masking." In this work, we analyze the effect of label smoothing on AT as one of the potential causes of gradient masking. We then develop a guided mechanism to avoid local minima during attack optimization, leading to a novel attack dubbed guided projected gradient attack (G-PGA). Our attack approach is based on a "match and deceive" loss that finds optimal adversarial directions through guidance from a surrogate model. Our modified attack does not require random restarts a large number of attack iterations or a search for optimal step size. Furthermore, our proposed G-PGA is generic, thus it can be combined with an ensemble attack strategy as we demonstrate in the case of auto-attack, leading to efficiency and convergence speed improvements. More than an effective attack, G-PGA can be used as a diagnostic tool to reveal elusive robustness due to gradient masking in adversarial defenses.

The success of deep learning based face recognition systems has given rise to serious privacy concerns due to their ability to enable unauthorized tracking of users in the digital world. Existing methods for enhancing privacy fail to generate “naturalistic” images that can protect facial privacy without compromising user experience. We propose a novel two-step approach for facial privacy protection that relies on finding adversarial latent codes in the low- dimensional manifold of a pretrained generative model. The first step inverts the given face image into the latent space and finetunes the generative model to achieve an accurate reconstruction of the given image from its latent code. This step produces a good initialization, aiding the generation of high-quality faces that resemble the given identity. Subsequently, user-defined makeup text prompts and identity- preserving regularization are used to guide the search for adversarial codes in the latent space. Extensive experiments demonstrate that faces generated by our approach have stronger black-box transferability with an absolute gain of 12.06% over the state-of-the-art facial privacy protection approach under the face verification task. Finally, we demonstrate the effectiveness of the proposed approach for commercial face recognition systems. Our code is available at https://github.com/fahadshamshad/Clip2Protect.

Adopting contrastive image-text pretrained models like CLIP towards video classification has gained attention due to its cost-effectiveness and competitive performance. However, recent works in this area face a trade-off. Finetuning the pretrained model to achieve strong supervised performance results in low zero-shot generalization. Similarly, freezing the backbone to retain zero-shot capability causes significant drop in supervised accuracy. Because of this, recent works in literature typically train separate models for supervised and zero-shot action recognition. In this work, we propose a multimodal prompt learning scheme that works to balance the supervised and zero-shot performance under a single unified training. Our prompting approach on the vision side caters for three aspects: 1) Global video-level prompts to model the data distribution; 2) Local frame-level prompts to provide per-frame discriminative conditioning; and 3) a summary prompt to extract a condensed video representation. Additionally, we define a prompting scheme on the text side to augment the textual context. Through this prompting scheme, we can achieve state-of-the-art zero-shot performance on Kinetics-600, HMDB51 and UCF101 while remaining competitive in the supervised setting. By keeping the pretrained backbone frozen, we optimize a much lower number of parameters and retain the existing general representation which helps achieve the strong zero-shot performance. Our codes/models will be released at https://github.com/TalalWasim/Vita-Clip..

Although existing semi-supervised learning models achieve remarkable success in learning with unannotated in-distribution data, they mostly fail to learn on unlabeled data sampled from novel semantic classes due to their closed-set assumption. In this work, we target a pragmatic but under-explored Generalized Novel Category Discovery (GNCD) setting. The GNCD setting aims to categorize unlabeled training data coming from known and novel classes by leveraging the information of partially labeled known classes. We propose a two-stage Contrastive Affinity Learning method with auxiliary visual Prompts, dubbed PromptCAL, to address this challenging problem. Our approach discovers reliable pairwise sample affinities to learn better semantic clustering of both known and novel classes for the class token and visual prompts. First, we propose a discriminative prompt regularization loss to reinforce semantic discriminativeness of prompt-adapted pre-trained vision transformer for refined affinity relationships. Besides, we propose contrastive affinity learning to calibrate semantic representations based on our iterative semi-supervised affinity graph generation method for semantically-enhanced supervision. Extensive experimental evaluation demonstrates that our PromptCAL method is more effective in discovering novel classes even with limited annotations and surpasses the current state-of-the-art on generic and fine-grained benchmarks (e.g., with nearly 11% gain on CUB-200, and 9% on ImageNet-100) on overall accuracy. Our code is available at https://github.com/sheng-eatamath/PromptCAL.

Prompt learning has emerged as an efficient alternative for fine-tuning foundational models, such as CLIP, for various downstream tasks. Conventionally trained using the task-specific objective, i.e., cross-entropy loss, prompts tend to overfit downstream data distributions and find it challenging to capture task-agnostic general features from the frozen CLIP. This leads to the loss of the model’s original generalization capability. To address this issue, our work introduces a self-regularization framework for prompting called PromptSRC (Prompting with Self-regulating Constraints). PromptSRC guides the prompts to optimize for both task-specific and task-agnostic general representations using a three-pronged approach by: (a) regulating prompted representations via mutual agreement maximization with the frozen model, (b) regulating with self-ensemble of prompts over the training trajectory to encode their complementary strengths, and (c) regulating with textual diversity to mitigate sample diversity imbalance with the visual branch. To the best of our knowledge, this is the first regularization framework for prompt learning that avoids overfitting by jointly attending to pre-trained model features, the training trajectory during prompting, and the textual diversity. PromptSRC explicitly steers the prompts to learn a representation space that maximizes performance on downstream tasks without compromising CLIP generalization. We perform extensive experiments on 4 benchmarks where PromptSRC overall performs favorably well compared to the existing methods. Our code and pre-trained models are publicly available at: https://github.com/muzairkhattak/PromptSRC.

Face anti-spoofing (FAS) or presentation attack detection is an essential component of face recognition systems deployed in security-critical applications. Existing FAS methods have poor generalizability to unseen spoof types, camera sensors, and environmental conditions. Recently, vision transformer (ViT) models have been shown to be effective for the FAS task due to their ability to capture long-range dependencies among image patches. However, adaptive modules or auxiliary loss functions are often required to adapt pre-trained ViT weights learned on large-scale datasets such as ImageNet. In this work, we first show that initializing ViTs with multimodal (e.g., CLIP) pre-trained weights improves generalizability for the FAS task, which is in line with the zero-shot transfer capabilities of vision-language pre-trained (VLP) models. We then propose a novel approach for robust cross-domain FAS by grounding visual representations with the help of natural language. Specifically, we show that aligning the image representation with an ensemble of class descriptions (based on natural language semantics) improves FAS generalizability in low-data regimes. Finally, we propose a multimodal contrastive learning strategy to boost feature generalization further and bridge the gap between source and target domains. Extensive experiments on three standard protocols demonstrate that our method significantly outperforms the state-of-the-art methods, achieving better zero-shot transfer performance than five-shot transfer of "adaptive ViTs". Code: https://github.com/koushiksrivats/FLIP

Recent video recognition models utilize Transformer models for long-range spatio-temporal context modeling. Video transformer designs are based on self-attention that can model global context at a high computational cost. In comparison, convolutional designs for videos offer an efficient alternative but lack long-range dependency modeling. Towards achieving the best of both designs, this work proposes Video-FocalNet, an effective and efficient architecture for video recognition that models both local and global contexts. Video-FocalNet is based on a spatiotemporal focal modulation architecture that reverses the interaction and aggregation steps of self-attention for better efficiency. Further, the aggregation step and the interaction step are both implemented using efficient convolution and element-wise multiplication operations that are computationally less expensive than their self-attention counterparts on video representations. We extensively explore the design space of focal modulation-based spatiotemporal context modeling and demonstrate our parallel spatial and temporal encoding design to be the optimal choice. Video-FocalNets perform favorably well against the state-of-the-art transformer-based models for video recognition on five large-scale datasets (Kinetics-400, Kinetics-600, SS-v2, Diving-48, and ActivityNet-1.3) at a lower computational cost. Our code/models are released at https://github.com/TalalWasim/Video-FocalNets.

The promising zero-shot generalization of vision-language models such as CLIP has led to their adoption using prompt learning for numerous downstream tasks. Previous works have shown test-time prompt tuning using entropy minimization to adapt text prompts for unseen domains. While effective, this overlooks the key cause for performance degradation to unseen domains -- distribution shift. In this work, we explicitly handle this problem by aligning the out-of-distribution (OOD) test sample statistics to those of the source data using prompt tuning. We use a single test sample to adapt multi-modal prompts at test time by minimizing the feature distribution shift to bridge the gap in the test domain. Evaluating against the domain generalization benchmark, our method improves zero-shot top-1 accuracy beyond existing prompt-learning techniques, with a 3.08% improvement over the baseline MaPLe. In cross-dataset generalization with unseen categories across 10 datasets, our method improves consistently across all datasets compared to the existing state-of-the-art. Our source code and models are available at [https://jameelhassan.github.io/promptalign](https://jameelhassan.github.io/promptalign)

Deep Convolution Neural Networks (CNNs) can easily be fooled by subtle, imperceptible changes to the input images. To address this vulnerability, adversarial training creates perturbation patterns and includes them in the training set to robustify the model. In contrast to existing adversarial training methods that only use class-boundary information (e.g., using a cross-entropy loss), we propose to exploit additional information from the feature space to craft stronger adversaries that are in turn used to learn a robust model. Specifically, we use the style and content information of the target sample from another class, alongside its class-boundary information to create adversarial perturbations. We apply our proposed multi-task objective in a deeply supervised manner, extracting multi-scale feature knowledge to create maximally separating adversaries. Subsequently, we propose a max-margin adversarial training approach that minimizes the distance between source image and its adversary and maximizes the distance between the adversary and the target image. Our adversarial training approach demonstrates strong robustness compared to state-of-the-art defenses, generalizes well to naturally occurring corruptions and data distributional shifts, and retains the model's accuracy on clean examples.

In this paper, we propose self-supervised training for video transformers using unlabeled video data. From a given video, we create local and global spatiotemporal views with varying spatial sizes and frame rates. Our self-supervised objective seeks to match the features of these different views representing the same video, to be invariant to spatiotemporal variations in actions. To the best of our knowledge, the proposed approach is the first to alleviate the dependency on negative samples or dedicated memory banks in Self-supervised Video Transformer (SVT). Further, owing to the flexibility of Transformer models, SVT supports slow-fast video processing within a single architecture using dynamically adjusted positional encoding and supports longterm relationship modeling along spatiotemporal dimensions. Our approach performs well on four action recognition benchmarks (Kinetics-400, UCF-101, HMDB-51, and SSv2) and converges faster with small batch sizes. Code is available at: https://git.io/J1juJ.

While the untargeted black-box transferability of adversarial perturbations has been extensively studied before, changing an unseen model’s decisions to a specific ‘targeted’ class remains a challenging feat. In this paper, we propose a new generative approach for highly transferable targeted perturbations (TTP). We note that the existing methods are less suitable for this task due to their reliance on class-boundary information that changes from one model to another, thus reducing transferability. In contrast, our approach matches the perturbed image ‘distribution’ with that of the target class, leading to high targeted transferability rates. To this end, we propose a new objective function that not only aligns the global distributions of source and target images, but also matches the local neighbourhood structure between the two domains. Based on the proposed objective, we train a generator function that can adaptively synthesize perturbations specific to a given input. Our generative approach is in-dependent of the source or target domain labels, while consistently performs well against state-of-the-art methods on a wide range of attack settings. As an example, we achieve 32.63% target transferability from (an adversarially weak) VGG19BNto (a strong) WideResNet on ImageNet val. set, which is 4× higher than the previous best generative attack and 16× better than instance-specific iterative attack. Code is available at: https://github.com/Muzammal-Naseer/TTP.

Deep neural networks have achieved remarkable performance on a range of classification tasks, with softmax crossentropy (CE) loss emerging as the de-facto objective function. The CE loss encourages features of a class to have a higher projection score on the true class-vector compared to the negative classes. However, this is a relative constraint and does not explicitly force different class features to be well-separated. Motivated by the observation that ground-truth class representations in CE loss are orthogonal (one-hot encoded vectors), we develop a novel loss function termed ‘Orthogonal Projection Loss' (OPL) which imposes orthogonality in the feature space. OPL augments the properties of CE loss and directly enforces inter-class separation alongside intra-class clustering in the feature space through orthogonality constraints on the mini-batch level. As compared to other alternatives of CE, OPL offers unique advantages e.g., no additional learnable parameters, does not require careful negative mining and is not sensitive to the batch size. Given the plug-and-play nature of OPL, we evaluate it on a diverse range of tasks including image recognition (CIFAR-100), large-scale classification (ImageNet), domain generalization (PACS) and few-shot learning (mini-ImageNet, CIFAR-FS, tiered-ImageNet and Meta-dataset) and demonstrate its effectiveness across the board. Furthermore, OPL offers better robustness against practical nuisances such as adversarial attacks and label noise. Code is available at: https://github.com/kahnchana/opl.

Vision transformers (ViT) have demonstrated impressive performance across numerous machine vision tasks. These models are based on multi-head self-attention mechanisms that can flexibly attend to a sequence of image patches to encode contextual cues. An important question is how such flexibility (in attending image-wide context conditioned on a given patch) can facilitate handling nuisances in natural images e.g., severe occlusions, domain shifts, spatial permutations, adversarial and natural perturbations. We systematically study this question via an extensive set of experiments encompassing three ViT families and provide comparisons with a high-performing convolutional neural network (CNN). We show and analyze the following intriguing properties of ViT: (a)Transformers are highly robust to severe occlusions, perturbations and domain shifts, e.g., retain as high as 60% top-1 accuracy on ImageNet even after randomly occluding 80% of the image content. (b)The robustness towards occlusions is not due to texture bias, instead we show that ViTs are significantly less biased towards local textures, compared to CNNs. When properly trained to encode shape-based features, ViTs demonstrate shape recognition capability comparable to that of human visual system, previously unmatched in the literature. (c)Using ViTs to encode shape representation leads to an interesting consequence of accurate semantic segmentation without pixel-level supervision. (d)Off-the-shelf features from a single ViT model can be combined to create a feature ensemble, leading to high accuracy rates across a range of classification datasets in both traditional and few-shot learning paradigms. We show effective features of ViTs are due to flexible and dynamic receptive fields possible via self-attention mechanisms. Our code will be publicly released.

Adversarial examples can cause catastrophic mistakes in Deep Neural Network (DNNs) based vision systems e.g., for classification, segmentation and object detection. The vulnerability of DNNs against such attacks can prove a major roadblock towards their real-world deployment. Transferability of adversarial examples demand generalizable defenses that can provide cross-task protection. Adversarial training that enhances robustness by modifying target model's parameters lacks such generalizability. On the other hand, different input processing based defenses fall short in the face of continuously evolving attacks. In this paper, we take the first step to combine the benefits of both approaches and propose a self-supervised adversarial training mechanism in the input space. By design, our defense is a generalizable approach and provides significant robustness against the unseen adversarial attacks (e.g. by reducing the success rate of translation-invariant ensemble attack from 82.6% to 31.9% in comparison to previous stateof-the-art). It can be deployed as a plug-and-play solution to protect a variety of vision systems, as we demonstrate for the case of classification, segmentation and detection. Code is available at: https://github.com/ Muzammal-Naseer/NRP.

Adversarial examples reveal the blind spots of deep neural networks (DNNs) and represent a major concern for security-critical applications. The transferability of adversarial examples makes real-world attacks possible in black-box settings, where the attacker is forbidden to access the internal parameters of the model. The underlying assumption in most adversary generation methods, whether learning an instance-specific or an instance-agnostic perturbation, is the direct or indirect reliance on the original domain-specific data distribution. In this work, for the first time, we demonstrate the existence of domain-invariant adversaries, thereby showing common adversarial space among different datasets and models. To this end, we propose a framework capable of launching highly transferable attacks that crafts adversarial patterns to mislead networks trained on wholly different domains. For instance, an adversarial function learned on Paintings, Cartoons or Medical images can successfully perturb ImageNet samples to fool the classifier, with success rates as high as $\sim$99\% ($\ell_{\infty} \le 10$). The core of our proposed adversarial function is a generative network that is trained using a relativistic supervisory signal that enables domain-invariant perturbations. Our approach sets the new state-of-the-art for fooling rates, both under the white-box and black-box scenarios. Furthermore, despite being an instance-agnostic perturbation function, our attack outperforms the conventionally much stronger instance-specific attack methods.

Deep neural networks (DNNs) have shown vulnerability to adversarial attacks, i.e., carefully perturbed inputs designed to mislead the network at inference time. Recently introduced localized attacks, Localized and Visible Adversarial Noise (LaVAN) and Adversarial patch, pose a new challenge to deep learning security by adding adversarial noise only within a specific region without affecting the salient objects in an image. Driven by the observation that such attacks introduce concentrated high-frequency changes at a particular image location, we have developed an effective method to estimate noise location in gradient domain and transform those high activation regions caused by adversarial noise in image domain while having minimal effect on the salient object that is important for correct classification. Our proposed Local Gradients Smoothing (LGS) scheme achieves this by regularizing gradients in the estimated noisy region before feeding the image to DNN for inference. We have shown the effectiveness of our method in comparison to other defense methods including Digital Watermarking, JPEG compression, Total Variance Minimization (TVM) and Feature squeezing on ImageNet dataset. In addition, we systematically study the robustness of the proposed defense mechanism against Back Pass Differentiable Approximation (BPDA), a state of the art attack recently developed to break defenses that transform an input sample to minimize the adversarial effect. Compared to other defense mechanisms, LGS is by far the most resistant to BPDA in localized adversarial attack setting.