Farid Boussaid

Aref Miri Rekavandi, Shima Rashidi, Farid Boussaid et al.

Transformers have rapidly gained popularity in computer vision, especially in the field of object recognition and detection. Upon examining the outcomes of state-of-the-art object detection methods, we noticed that transformers consistently outperformed well-established CNN-based detectors in almost every video or image dataset. While transformer-based approaches remain at the forefront of small object detection (SOD) techniques, this paper aims to explore the performance benefits offered by such extensive networks and identify potential reasons for their SOD superiority. Small objects have been identified as one of the most challenging object types in detection frameworks due to their low visibility. We aim to investigate potential strategies that could enhance transformers' performance in SOD. This survey presents a taxonomy of over 60 research studies on developed transformers for the task of SOD, spanning the years 2020 to 2023. These studies encompass a variety of detection applications, including small object detection in generic images, aerial images, medical images, active millimeter images, underwater images, and videos. We also compile and present a list of 12 large-scale datasets suitable for SOD that were overlooked in previous studies and compare the performance of the reviewed studies using popular metrics such as mean Average Precision (mAP), Frames Per Second (FPS), number of parameters, and more. Researchers can keep track of newer studies on our web page, which is available at \url{https://github.com/arekavandi/Transformer-SOD}.

Yanan Liu, Anqi Zhu, Jingmin Zhu et al.

Skeleton-based action recognition aims to understand human behaviors from body joint sequences and is especially challenging in the one-shot setting, where only a single labeled exemplar is available for each novel action. A key challenge is learning representations that capture the hierarchical and compositional structure of human motion while aligning effectively with high-level action semantics under extreme data scarcity. Existing approaches, largely based on Euclidean embeddings and low-level motion cues, struggle to model the tree-like organization of skeleton data, limiting cross-modal alignment and generalization to unseen action categories. We propose SkelHCC, a unified skeleton hyperbolic CLIP-driven cache adaptation framework for one-shot skeleton-based action recognition. SkelHCC introduces an Explicitly Hierarchical Hyperbolic CLIP (EH-HCLIP) module that embeds skeleton sequences and action language into a shared hyperbolic space. By leveraging the negative curvature and exponential volume growth of hyperbolic geometry, EH-HCLIP naturally encodes the joint-part-body hierarchy of human anatomy and yields structurally consistent cross-modal representations. To support efficient one-shot adaptation, SkelHCC further integrates a training-free LLM-guided Multi-granularity Voting Cache (LMV-Cache) for context-aware inference. Experiments on NTU RGB+D 60, NTU RGB+D 120, and PKU-MMD demonstrate that SkelHCC consistently outperforms state-of-the-art methods.

Lin Wu, Deyin Liu, Wenying Zhang et al. · ibm-research

Person re-identification (re-ID) is of great importance to video surveillance systems by estimating the similarity between a pair of cross-camera person shorts. Current methods for estimating such similarity require a large number of labeled samples for supervised training. In this paper, we present a pseudo-pair based self-similarity learning approach for unsupervised person re-ID without human annotations. Unlike conventional unsupervised re-ID methods that use pseudo labels based on global clustering, we construct patch surrogate classes as initial supervision, and propose to assign pseudo labels to images through the pairwise gradient-guided similarity separation. This can cluster images in pseudo pairs, and the pseudos can be updated during training. Based on pseudo pairs, we propose to improve the generalization of similarity function via a novel self-similarity learning:it learns local discriminative features from individual images via intra-similarity, and discovers the patch correspondence across images via inter-similarity. The intra-similarity learning is based on channel attention to detect diverse local features from an image. The inter-similarity learning employs a deformable convolution with a non-local block to align patches for cross-image similarity. Experimental results on several re-ID benchmark datasets demonstrate the superiority of the proposed method over the state-of-the-arts.

Yanbin Liu, Girish Dwivedi, Farid Boussaid et al.

Generative models such as generative adversarial networks and autoencoders have gained a great deal of attention in the medical field due to their excellent data generation capability. This paper provides a comprehensive survey of generative models for three-dimensional (3D) volumes, focusing on the brain and heart. A new and elaborate taxonomy of unconditional and conditional generative models is proposed to cover diverse medical tasks for the brain and heart: unconditional synthesis, classification, conditional synthesis, segmentation, denoising, detection, and registration. We provide relevant background, examine each task and also suggest potential future directions. A list of the latest publications will be updated on Github to keep up with the rapid influx of papers at https://github.com/csyanbin/3D-Medical-Generative-Survey.

Lian Xu, Wanli Ouyang, Mohammed Bennamoun et al.

This paper proposes a new transformer-based framework to learn class-specific object localization maps as pseudo labels for weakly supervised semantic segmentation (WSSS). Inspired by the fact that the attended regions of the one-class token in the standard vision transformer can be leveraged to form a class-agnostic localization map, we investigate if the transformer model can also effectively capture class-specific attention for more discriminative object localization by learning multiple class tokens within the transformer. To this end, we propose a Multi-class Token Transformer, termed as MCTformer, which uses multiple class tokens to learn interactions between the class tokens and the patch tokens. The proposed MCTformer can successfully produce class-discriminative object localization maps from class-to-patch attentions corresponding to different class tokens. We also propose to use a patch-level pairwise affinity, which is extracted from the patch-to-patch transformer attention, to further refine the localization maps. Moreover, the proposed framework is shown to fully complement the Class Activation Mapping (CAM) method, leading to remarkably superior WSSS results on the PASCAL VOC and MS COCO datasets. These results underline the importance of the class token for WSSS.

Lian Xu, Mohammed Bennamoun, Farid Boussaid et al.

This paper proposes a novel transformer-based framework that aims to enhance weakly supervised semantic segmentation (WSSS) by generating accurate class-specific object localization maps as pseudo labels. Building upon the observation that the attended regions of the one-class token in the standard vision transformer can contribute to a class-agnostic localization map, we explore the potential of the transformer model to capture class-specific attention for class-discriminative object localization by learning multiple class tokens. We introduce a Multi-Class Token transformer, which incorporates multiple class tokens to enable class-aware interactions with the patch tokens. To achieve this, we devise a class-aware training strategy that establishes a one-to-one correspondence between the output class tokens and the ground-truth class labels. Moreover, a Contrastive-Class-Token (CCT) module is proposed to enhance the learning of discriminative class tokens, enabling the model to better capture the unique characteristics and properties of each class. As a result, class-discriminative object localization maps can be effectively generated by leveraging the class-to-patch attentions associated with different class tokens. To further refine these localization maps, we propose the utilization of patch-level pairwise affinity derived from the patch-to-patch transformer attention. Furthermore, the proposed framework seamlessly complements the Class Activation Mapping (CAM) method, resulting in significantly improved WSSS performance on the PASCAL VOC 2012 and MS COCO 2014 datasets. These results underline the importance of the class token for WSSS.

Aref Miri Rekavandi, Lian Xu, Farid Boussaid et al.

Small object detection (SOD) in optical images and videos is a challenging problem that even state-of-the-art generic object detection methods fail to accurately localize and identify such objects. Typically, small objects appear in real-world due to large camera-object distance. Because small objects occupy only a small area in the input image (e.g., less than 10%), the information extracted from such a small area is not always rich enough to support decision making. Multidisciplinary strategies are being developed by researchers working at the interface of deep learning and computer vision to enhance the performance of SOD deep learning based methods. In this paper, we provide a comprehensive review of over 160 research papers published between 2017 and 2022 in order to survey this growing subject. This paper summarizes the existing literature and provide a taxonomy that illustrates the broad picture of current research. We investigate how to improve the performance of small object detection in maritime environments, where increasing performance is critical. By establishing a connection between generic and maritime SOD research, future directions have been identified. In addition, the popular datasets that have been used for SOD for generic and maritime applications are discussed, and also well-known evaluation metrics for the state-of-the-art methods on some of the datasets are provided.

Mohammed Hassanin, Abdelwahed Khamiss, Mohammed Bennamoun et al.

3D human pose estimation can be handled by encoding the geometric dependencies between the body parts and enforcing the kinematic constraints. Recently, Transformer has been adopted to encode the long-range dependencies between the joints in the spatial and temporal domains. While they had shown excellence in long-range dependencies, studies have noted the need for improving the locality of vision Transformers. In this direction, we propose a novel pose estimation Transformer featuring rich representations of body joints critical for capturing subtle changes across frames (i.e., inter-feature representation). Specifically, through two novel interaction modules; Cross-Joint Interaction and Cross-Frame Interaction, the model explicitly encodes the local and global dependencies between the body joints. The proposed architecture achieved state-of-the-art performance on two popular 3D human pose estimation datasets, Human3.6 and MPI-INF-3DHP. In particular, our proposed CrossFormer method boosts performance by 0.9% and 0.3%, compared to the closest counterpart, PoseFormer, using the detected 2D poses and ground-truth settings respectively.

Lin Wu, Lingqiao Liu, Yang Wang et al.

The cross-resolution person re-identification (CRReID) problem aims to match low-resolution (LR) query identity images against high resolution (HR) gallery images. It is a challenging and practical problem since the query images often suffer from resolution degradation due to the different capturing conditions from real-world cameras. To address this problem, state-of-the-art (SOTA) solutions either learn the resolution-invariant representation or adopt super-resolution (SR) module to recover the missing information from the LR query. This paper explores an alternative SR-free paradigm to directly compare HR and LR images via a dynamic metric, which is adaptive to the resolution of a query image. We realize this idea by learning resolution-adaptive representations for cross-resolution comparison. Specifically, we propose two resolution-adaptive mechanisms. The first one disentangles the resolution-specific information into different sub-vectors in the penultimate layer of the deep neural networks, and thus creates a varying-length representation. To better extract resolution-dependent information, we further propose to learn resolution-adaptive masks for intermediate residual feature blocks. A novel progressive learning strategy is proposed to train those masks properly. These two mechanisms are combined to boost the performance of CRReID. Experimental results show that the proposed method is superior to existing approaches and achieves SOTA performance on multiple CRReID benchmarks.

Yanbin Liu, Girish Dwivedi, Farid Boussaid et al.

The generation of three-dimensional (3D) medical images has great application potential since it takes into account the 3D anatomical structure. Two problems prevent effective training of a 3D medical generative model: (1) 3D medical images are expensive to acquire and annotate, resulting in an insufficient number of training images, and (2) a large number of parameters are involved in 3D convolution. Methods: We propose a novel GAN model called 3D Split&Shuffle-GAN. To address the 3D data scarcity issue, we first pre-train a two-dimensional (2D) GAN model using abundant image slices and inflate the 2D convolution weights to improve the initialization of the 3D GAN. Novel 3D network architectures are proposed for both the generator and discriminator of the GAN model to significantly reduce the number of parameters while maintaining the quality of image generation. Several weight inflation strategies and parameter-efficient 3D architectures are investigated. Results: Experiments on both heart (Stanford AIMI Coronary Calcium) and brain (Alzheimer's Disease Neuroimaging Initiative) datasets show that our method leads to improved 3D image generation quality (14.7 improvements on Fréchet inception distance) with significantly fewer parameters (only 48.5% of the baseline method). Conclusions: We built a parameter-efficient 3D medical image generation model. Due to the efficiency and effectiveness, it has the potential to generate high-quality 3D brain and heart images for real use cases.

Laurent Jospin, Allen Antony, Lian Xu et al.

In stereo vision, self-similar or bland regions can make it difficult to match patches between two images. Active stereo-based methods mitigate this problem by projecting a pseudo-random pattern on the scene so that each patch of an image pair can be identified without ambiguity. However, the projected pattern significantly alters the appearance of the image. If this pattern acts as a form of adversarial noise, it could negatively impact the performance of deep learning-based methods, which are now the de-facto standard for dense stereo vision. In this paper, we propose the Active-Passive SimStereo dataset and a corresponding benchmark to evaluate the performance gap between passive and active stereo images for stereo matching algorithms. Using the proposed benchmark and an additional ablation study, we show that the feature extraction and matching modules of a selection of twenty selected deep learning-based stereo matching methods generalize to active stereo without a problem. However, the disparity refinement modules of three of the twenty architectures (ACVNet, CascadeStereo, and StereoNet) are negatively affected by the active stereo patterns due to their reliance on the appearance of the input images.

Ashkan Taghipour, Morteza Ghahremani, Mohammed Bennamoun et al.

This paper investigates the role of CLIP image embeddings within the Stable Video Diffusion (SVD) framework, focusing on their impact on video generation quality and computational efficiency. Our findings indicate that CLIP embeddings, while crucial for aesthetic quality, do not significantly contribute towards the subject and background consistency of video outputs. Moreover, the computationally expensive cross-attention mechanism can be effectively replaced by a simpler linear layer. This layer is computed only once at the first diffusion inference step, and its output is then cached and reused throughout the inference process, thereby enhancing efficiency while maintaining high-quality outputs. Building on these insights, we introduce the VCUT, a training-free approach optimized for efficiency within the SVD architecture. VCUT eliminates temporal cross-attention and replaces spatial cross-attention with a one-time computed linear layer, significantly reducing computational load. The implementation of VCUT leads to a reduction of up to 322T Multiple-Accumulate Operations (MACs) per video and a decrease in model parameters by up to 50M, achieving a 20% reduction in latency compared to the baseline. Our approach demonstrates that conditioning during the Semantic Binding stage is sufficient, eliminating the need for continuous computation across all inference steps and setting a new standard for efficient video generation.

Jingmin Zhu, Anqi Zhu, James Bailey et al.

Zero-shot skeleton-based action recognition (ZS-SAR) is fundamentally constrained by prevailing approaches that rely on aligning skeleton features with static, class-level semantics. This coarse-grained alignment fails to bridge the domain shift between seen and unseen classes, thereby impeding the effective transfer of fine-grained visual knowledge. To address these limitations, we introduce \textbf{DynaPURLS}, a unified framework that establishes robust, multi-scale visual-semantic correspondences and dynamically refines them at inference time to enhance generalization. Our framework leverages a large language model to generate hierarchical textual descriptions that encompass both global movements and local body-part dynamics. Concurrently, an adaptive partitioning module produces fine-grained visual representations by semantically grouping skeleton joints. To fortify this fine-grained alignment against the train-test domain shift, DynaPURLS incorporates a dynamic refinement module. During inference, this module adapts textual features to the incoming visual stream via a lightweight learnable projection. This refinement process is stabilized by a confidence-aware, class-balanced memory bank, which mitigates error propagation from noisy pseudo-labels. Extensive experiments on three large-scale benchmark datasets, including NTU RGB+D 60/120 and PKU-MMD, demonstrate that DynaPURLS significantly outperforms prior art, setting new state-of-the-art records. The source code is made publicly available at https://github.com/Alchemist0754/DynaPURLS

Yiheng Lyu, Lian Xu, Mohammed Bennamoun et al.

Weakly supervised semantic segmentation offers a label-efficient solution to train segmentation models for volumetric medical imaging. However, existing approaches often rely on 2D encoders that neglect the inherent volumetric nature of the data. We propose TranSamba, a hybrid Transformer-Mamba architecture designed to capture 3D context for weakly supervised volumetric medical segmentation. TranSamba augments a standard Vision Transformer backbone with Cross-Plane Mamba blocks, which leverage the linear complexity of state space models for efficient information exchange across neighboring slices. The information exchange enhances the pairwise self-attention within slices computed by the Transformer blocks, directly contributing to the attention maps for object localization. TranSamba achieves effective volumetric modeling with time complexity that scales linearly with the input volume depth and maintains constant memory usage for batch processing. Extensive experiments on three datasets demonstrate that TranSamba establishes new state-of-the-art performance, consistently outperforming existing methods across diverse modalities and pathologies. Our source code and trained models are openly accessible at: https://github.com/YihengLyu/TranSamba.

Deyin Liu, Lin Wu, Haifeng Zhao et al.

Deep neural networks (DNNs) are known to be vulnerable to adversarial examples that are crafted with imperceptible perturbations, i.e., a small change in an input image can induce a mis-classification, and thus threatens the reliability of deep learning based deployment systems. Adversarial training (AT) is often adopted to improve robustness through training a mixture of corrupted and clean data. However, most of AT based methods are ineffective in dealing with transferred adversarial examples which are generated to fool a wide spectrum of defense models, and thus cannot satisfy the generalization requirement raised in real-world scenarios. Moreover, adversarially training a defense model in general cannot produce interpretable predictions towards the inputs with perturbations, whilst a highly interpretable robust model is required by different domain experts to understand the behaviour of a DNN. In this work, we propose a novel approach based on Jacobian norm and Selective Input Gradient Regularization (J-SIGR), which suggests the linearized robustness through Jacobian normalization and also regularizes the perturbation-based saliency maps to imitate the model's interpretable predictions. As such, we achieve both the improved defense and high interpretability of DNNs. Finally, we evaluate our method across different architectures against powerful adversarial attacks. Experiments demonstrate that the proposed J-SIGR confers improved robustness against transferred adversarial attacks, and we also show that the predictions from the neural network are easy to interpret.

Hao Tang, Aref Miri Rekavandi, Dharjinder Rooprai et al.

This study investigates the effectiveness of Explainable Artificial Intelligence (XAI) techniques in predicting suicide risks and identifying the dominant causes for such behaviours. Data augmentation techniques and ML models are utilized to predict the associated risk. Furthermore, SHapley Additive exPlanations (SHAP) and correlation analysis are used to rank the importance of variables in predictions. Experimental results indicate that Decision Tree (DT), Random Forest (RF) and eXtreme Gradient Boosting (XGBoost) models achieve the best results while DT has the best performance with an accuracy of 95:23% and an Area Under Curve (AUC) of 0.95. As per SHAP results, anger problems, depression, and social isolation are the leading variables in predicting the risk of suicide, and patients with good incomes, respected occupations, and university education have the least risk. Results demonstrate the effectiveness of machine learning and XAI framework for suicide risk prediction, and they can assist psychiatrists in understanding complex human behaviours and can also assist in reliable clinical decision-making.

Laurent Valentin Jospin, Hamid Laga, Farid Boussaid et al.

A major focus of recent developments in stereo vision has been on how to obtain accurate dense disparity maps in passive stereo vision. Active vision systems enable more accurate estimations of dense disparity compared to passive stereo. However, subpixel-accurate disparity estimation remains an open problem that has received little attention. In this paper, we propose a new learning strategy to train neural networks to estimate high-quality subpixel disparity maps for semi-dense active stereo vision. The key insight is that neural networks can double their accuracy if they are able to jointly learn how to refine the disparity map while invalidating the pixels where there is insufficient information to correct the disparity estimate. Our approach is based on Bayesian modeling where validated and invalidated pixels are defined by their stochastic properties, allowing the model to learn how to choose by itself which pixels are worth its attention. Using active stereo datasets such as Active-Passive SimStereo, we demonstrate that the proposed method outperforms the current state-of-the-art active stereo models. We also demonstrate that the proposed approach compares favorably with state-of-the-art passive stereo models on the Middlebury dataset.

Tahmina Khanam, Hamid Laga, Mohammed Bennamoun et al.

We propose the first comprehensive approach for modeling and analyzing the spatiotemporal shape variability in tree-like 4D objects, i.e., 3D objects whose shapes bend, stretch, and change in their branching structure over time as they deform, grow, and interact with their environment. Our key contribution is the representation of tree-like 3D shapes using Square Root Velocity Function Trees (SRVFT). By solving the spatial registration in the SRVFT space, which is equipped with an L2 metric, 4D tree-shaped structures become time-parameterized trajectories in this space. This reduces the problem of modeling and analyzing 4D tree-like shapes to that of modeling and analyzing elastic trajectories in the SRVFT space, where elasticity refers to time warping. In this paper, we propose a novel mathematical representation of the shape space of such trajectories, a Riemannian metric on that space, and computational tools for fast and accurate spatiotemporal registration and geodesics computation between 4D tree-shaped structures. Leveraging these building blocks, we develop a full framework for modelling the spatiotemporal variability using statistical models and generating novel 4D tree-like structures from a set of exemplars. We demonstrate and validate the proposed framework using real 4D plant data.

A S M Sharifuzzaman Sagar, Mohammed Bennamoun, Farid Boussaid et al.

In multimodal misinformation, deception usually arises not just from pixel-level manipulations in an image, but from the semantic and contextual claim jointly expressed by the image-text pair. Yet most deepfake detectors, engineered to detect pixel-level forgeries, do not account for claim-level meaning, despite their growing integration in automated fact-checking (AFC) pipelines. This raises a central scientific and practical question: Do pixel-level detectors contribute useful signal for verifying image-text claims, or do they instead introduce misleading authenticity priors that undermine evidence-based reasoning? We provide the first systematic analysis of deepfake detectors in the context of multimodal misinformation detection. Using two complementary benchmarks, MMFakeBench and DGM4, we evaluate: (1) state-of-the-art image-only deepfake detectors, (2) an evidence-driven fact-checking system that performs tool-guided retrieval via Monte Carlo Tree Search (MCTS) and engages in deliberative inference through Multi-Agent Debate (MAD), and (3) a hybrid fact-checking system that injects detector outputs as auxiliary evidence. Results across both benchmark datasets show that deepfake detectors offer limited standalone value, achieving F1 scores in the range of 0.26-0.53 on MMFakeBench and 0.33-0.49 on DGM4, and that incorporating their predictions into fact-checking pipelines consistently reduces performance by 0.04-0.08 F1 due to non-causal authenticity assumptions. In contrast, the evidence-centric fact-checking system achieves the highest performance, reaching F1 scores of approximately 0.81 on MMFakeBench and 0.55 on DGM4. Overall, our findings demonstrate that multimodal claim verification is driven primarily by semantic understanding and external evidence, and that pixel-level artifact signals do not reliably enhance reasoning over real-world image-text misinformation.

Ashkan Taghipour, Morteza Ghahremani, Mohammed Bennamoun et al.

While latent diffusion models (LDMs) excel at creating imaginative images, they often lack precision in semantic fidelity and spatial control over where objects are generated. To address these deficiencies, we introduce the Box-it-to-Bind-it (B2B) module - a novel, training-free approach for improving spatial control and semantic accuracy in text-to-image (T2I) diffusion models. B2B targets three key challenges in T2I: catastrophic neglect, attribute binding, and layout guidance. The process encompasses two main steps: i) Object generation, which adjusts the latent encoding to guarantee object generation and directs it within specified bounding boxes, and ii) attribute binding, guaranteeing that generated objects adhere to their specified attributes in the prompt. B2B is designed as a compatible plug-and-play module for existing T2I models, markedly enhancing model performance in addressing the key challenges. We evaluate our technique using the established CompBench and TIFA score benchmarks, demonstrating significant performance improvements compared to existing methods. The source code will be made publicly available at https://github.com/nextaistudio/BoxIt2BindIt.

Xian Zhang, Zexi Wu, Zinuo Li et al.

Understanding long-form videos remains a significant challenge for vision--language models (VLMs) due to their extensive temporal length and high information density. Most current multimodal large language models (MLLMs) rely on uniform sampling, which often overlooks critical moments, leading to incorrect responses to queries. In parallel, many keyframe selection approaches impose rigid temporal spacing: once a frame is chosen, an exclusion window suppresses adjacent timestamps to reduce redundancy. While effective at limiting overlap, this strategy frequently misses short, fine-grained cues near important events. Other methods instead emphasize visual diversity but neglect query relevance. We propose AdaRD-Key, a training-free keyframe sampling module for query-driven long-form video understanding. AdaRD-Key maximizes a unified Relevance--Diversity Max-Volume (RD-MV) objective, combining a query-conditioned relevance score with a log-determinant diversity component to yield informative yet non-redundant frames. To handle broad queries with weak alignment to the video, AdaRD-Key employs a lightweight relevance-aware gating mechanism; when the relevance distribution indicates weak alignment, the method seamlessly shifts into a diversity-only mode, enhancing coverage without additional supervision. Our pipeline is training-free, computationally efficient (running in real time on a single GPU), and compatible with existing VLMs in a plug-and-play manner. Extensive experiments on LongVideoBench and Video-MME demonstrate state-of-the-art performance, particularly on long-form videos. Code available at https://github.com/Xian867/AdaRD-Key.

Lian Xu, Mohammed Bennamoun, Farid Boussaid et al.

Most existing weakly supervised semantic segmentation (WSSS) methods rely on Class Activation Mapping (CAM) to extract coarse class-specific localization maps using image-level labels. Prior works have commonly used an off-line heuristic thresholding process that combines the CAM maps with off-the-shelf saliency maps produced by a general pre-trained saliency model to produce more accurate pseudo-segmentation labels. We propose AuxSegNet+, a weakly supervised auxiliary learning framework to explore the rich information from these saliency maps and the significant inter-task correlation between saliency detection and semantic segmentation. In the proposed AuxSegNet+, saliency detection and multi-label image classification are used as auxiliary tasks to improve the primary task of semantic segmentation with only image-level ground-truth labels. We also propose a cross-task affinity learning mechanism to learn pixel-level affinities from the saliency and segmentation feature maps. In particular, we propose a cross-task dual-affinity learning module to learn both pairwise and unary affinities, which are used to enhance the task-specific features and predictions by aggregating both query-dependent and query-independent global context for both saliency detection and semantic segmentation. The learned cross-task pairwise affinity can also be used to refine and propagate CAM maps to provide better pseudo labels for both tasks. Iterative improvement of segmentation performance is enabled by cross-task affinity learning and pseudo-label updating. Extensive experiments demonstrate the effectiveness of the proposed approach with new state-of-the-art WSSS results on the challenging PASCAL VOC and MS COCO benchmarks.

Ning Wang, Tieyue Wu, Naeha Sharif et al.

Zero-shot skeleton-based action recognition aims to recognize unseen actions by transferring knowledge from seen categories through semantic descriptions. Most existing methods typically align skeleton features with textual embeddings within a shared latent space. However, the absence of contextual cues, such as objects involved in the action, introduces an inherent gap between skeleton and semantic representations, making it difficult to distinguish visually similar actions. To address this, we propose SkeletonContext, a prompt-based framework that enriches skeletal motion representations with language-driven contextual semantics. Specifically, we introduce a Cross-Modal Context Prompt Module, which leverages a pretrained language model to reconstruct masked contextual prompts under guidance derived from LLMs. This design effectively transfers linguistic context to the skeleton encoder for instance-level semantic grounding and improved cross-modal alignment. In addition, a Key-Part Decoupling Module is incorporated to decouple motion-relevant joint features, ensuring robust action understanding even in the absence of explicit object interactions. Extensive experiments on multiple benchmarks demonstrate that SkeletonContext achieves state-of-the-art performance under both conventional and generalized zero-shot settings, validating its effectiveness in reasoning about context and distinguishing fine-grained, visually similar actions.

Awais Nizamani, Hamid Laga, Guanjin Wang et al.

We propose a novel framework for the statistical analysis of genus-zero 4D surfaces, i.e., 3D surfaces that deform and evolve over time. This problem is particularly challenging due to the arbitrary parameterizations of these surfaces and their varying deformation speeds, necessitating effective spatiotemporal registration. Traditionally, 4D surfaces are discretized, in space and time, before computing their spatiotemporal registrations, geodesics, and statistics. However, this approach may result in suboptimal solutions and, as we demonstrate in this paper, is not necessary. In contrast, we treat 4D surfaces as continuous functions in both space and time. We introduce Dynamic Spherical Neural Surfaces (D-SNS), an efficient smooth and continuous spatiotemporal representation for genus-0 4D surfaces. We then demonstrate how to perform core 4D shape analysis tasks such as spatiotemporal registration, geodesics computation, and mean 4D shape estimation, directly on these continuous representations without upfront discretization and meshing. By integrating neural representations with classical Riemannian geometry and statistical shape analysis techniques, we provide the building blocks for enabling full functional shape analysis. We demonstrate the efficiency of the framework on 4D human and face datasets. The source code and additional results are available at https://4d-dsns.github.io/DSNS/.

Zinuo Li, Yongxin Guo, Jun Liu et al.

Human understanding of video dynamics is typically grounded in a structured mental representation of entities, actions, and temporal relations, rather than relying solely on immediate deductive reasoning. In contrast, existing Video-LLMs largely depend on unstructured video reasoning, where critical visual evidence is embedded in verbose textual descriptions and temporal causality is often weakly modeled. This leads to inefficient processes and fragile causal inference. To bridge this cognitive gap, we propose constructing a compact representation of salient events and their causal relationships, which we name Structured Event Facts, prior to the reasoning stage. This structured prior serves as an explicit constraint to promote concise and causally grounded reasoning, while also making intermediate evidence easier to verify. To effectively train models on such structured facts, we introduce CausalFact-60K and a four-stage training pipeline comprising facts alignment, format warm-start, thinking warm-start, and reinforcement learning-based post-training. During RL stage, we find that this framework introduces competing objectives, as structural completeness and causal fidelity must be balanced against reasoning length, making it difficult to optimize. We address this challenge by formulating the optimization as a Multi-Objective Reinforcement Learning (MORL) problem and explicitly optimizing toward the Pareto-Frontier to balance these trade-offs. As a result, we introduce Factum-4B, which yields more reliable reasoning and delivers stronger performance on challenging video understanding tasks requiring fine-grained temporal inference.

Ashkan Taghipour, Morteza Ghahremani, Mohammed Bennamoun et al.

Image-to-video (I2V) generation seeks to produce realistic motion sequences from a single reference image. Although recent methods exhibit strong temporal consistency, they often struggle when dealing with complex, non-repetitive human movements, leading to unnatural deformations. To tackle this issue, we present LatentMove, a DiT-based framework specifically tailored for highly dynamic human animation. Our architecture incorporates a conditional control branch and learnable face/body tokens to preserve consistency as well as fine-grained details across frames. We introduce Complex-Human-Videos (CHV), a dataset featuring diverse, challenging human motions designed to benchmark the robustness of I2V systems. We also introduce two metrics to assess the flow and silhouette consistency of generated videos with their ground truth. Experimental results indicate that LatentMove substantially improves human animation quality--particularly when handling rapid, intricate movements--thereby pushing the boundaries of I2V generation. The code, the CHV dataset, and the evaluation metrics will be available at https://github.com/ --.

Zinuo Li, Xian Zhang, Yongxin Guo et al.

Humans naturally understand moments in a video by integrating visual and auditory cues. For example, localizing a scene in the video like "A scientist passionately speaks on wildlife conservation as dramatic orchestral music plays, with the audience nodding and applauding" requires simultaneous processing of visual, audio, and speech signals. However, existing models often struggle to effectively fuse and interpret audio information, limiting their capacity for comprehensive video temporal understanding. To address this, we present TriSense, a triple-modality large language model designed for holistic video temporal understanding through the integration of visual, audio, and speech modalities. Central to TriSense is a Query-Based Connector that adaptively reweights modality contributions based on the input query, enabling robust performance under modality dropout and allowing flexible combinations of available inputs. To support TriSense's multimodal capabilities, we introduce TriSense-2M, a high-quality dataset of over 2 million curated samples generated via an automated pipeline powered by fine-tuned LLMs. TriSense-2M includes long-form videos and diverse modality combinations, facilitating broad generalization. Extensive experiments across multiple benchmarks demonstrate the effectiveness of TriSense and its potential to advance multimodal video analysis. Code and dataset will be publicly released.

Chengyuan Zhang, Yilin Zhang, Lei Zhu et al.

This paper introduces a novel framework for unified incremental few-shot object detection (iFSOD) and instance segmentation (iFSIS) using the Transformer architecture. Our goal is to create an optimal solution for situations where only a few examples of novel object classes are available, with no access to training data for base or old classes, while maintaining high performance across both base and novel classes. To achieve this, We extend Mask-DINO into a two-stage incremental learning framework. Stage 1 focuses on optimizing the model using the base dataset, while Stage 2 involves fine-tuning the model on novel classes. Besides, we incorporate a classifier selection strategy that assigns appropriate classifiers to the encoder and decoder according to their distinct functions. Empirical evidence indicates that this approach effectively mitigates the over-fitting on novel classes learning. Furthermore, we implement knowledge distillation to prevent catastrophic forgetting of base classes. Comprehensive evaluations on the COCO and LVIS datasets for both iFSIS and iFSOD tasks demonstrate that our method significantly outperforms state-of-the-art approaches.

Ashkan Taghipour, Morteza Ghahremani, Zinuo Li et al.

Generating videos of complex human motions such as flips, cartwheels, and martial arts remains challenging for current video diffusion models. Text-only conditioning is temporally ambiguous for fine-grained motion control, while explicit pose-based controls, though effective, require users to provide complete skeleton sequences that are costly to produce for long and dynamic actions. We propose a two-stage cascaded framework that addresses both limitations. First, an autoregressive text-to-skeleton model generates 2D pose sequences from natural language descriptions by predicting each joint conditioned on previously generated poses. This design captures long-range temporal dependencies and inter-joint coordination required for complex motions. Second, a pose-conditioned video diffusion model synthesizes videos from a reference image and the generated skeleton sequence. It employs DINO-ALF (Adaptive Layer Fusion), a multi-level reference encoder that preserves appearance and clothing details under large pose changes and self-occlusions. To address the lack of publicly available datasets for complex human motion video generation, we introduce a Blender-based synthetic dataset containing 2,000 videos with diverse characters performing acrobatic and stunt-like motions. The dataset provides full control over appearance, motion, and environment. It fills an important gap because existing benchmarks significantly under-represent acrobatic motions while web-collected datasets raise copyright and privacy concerns. Experiments on our synthetic dataset and the Motion-X Fitness benchmark show that our text-to-skeleton model outperforms prior methods on FID, R-precision, and motion diversity. Our pose-to-video model also achieves the best results among all compared methods on VBench metrics for temporal consistency, motion smoothness, and subject preservation.

Ashkan Taghipour, Vahid Naghshin, Benjamin Southwell et al.

3D Gaussian Splatting (3DGS) enables fast, high-quality novel view synthesis but typically relies on densification followed by pruning to optimize the number of Gaussians. Existing mask-based pruning, such as MaskGS, regularizes the global mean of the mask, which is misaligned with the local per-pixel (per-ray) reconstruction loss that determines image quality along individual camera rays. This paper introduces SVR-GS, a spatially variant regularizer that renders a per-pixel spatial mask from each Gaussian's effective contribution along the ray, thereby applying sparsity pressure where it matters: on low-importance Gaussians. We explore three spatial-mask aggregation strategies, implement them in CUDA, and conduct a gradient analysis to motivate our final design. Extensive experiments on Tanks\&Temples, Deep Blending, and Mip-NeRF360 datasets demonstrate that, on average across the three datasets, the proposed SVR-GS reduces the number of Gaussians by 1.79\(\times\) compared to MaskGS and 5.63\(\times\) compared to 3DGS, while incurring only 0.50 dB and 0.40 dB PSNR drops, respectively. These gains translate into significantly smaller, faster, and more memory-efficient models, making them well-suited for real-time applications such as robotics, AR/VR, and mobile perception.

Laurent Valentin Jospin, Farid Boussaid, Hamid Laga et al.

Cost-based image patch matching is at the core of various techniques in computer vision, photogrammetry and remote sensing. When the subpixel disparity between the reference patch in the source and target images is required, either the cost function or the target image have to be interpolated. While cost-based interpolation is the easiest to implement, multiple works have shown that image based interpolation can increase the accuracy of the subpixel matching, but usually at the cost of expensive search procedures. This, however, is problematic, especially for very computation intensive applications such as stereo matching or optical flow computation. In this paper, we show that closed form formulae for subpixel disparity computation for the case of one dimensional matching, e.g., in the case of rectified stereo images where the search space is of one dimension, exists when using the standard NCC, SSD and SAD cost functions. We then demonstrate how to generalize the proposed formulae to the case of high dimensional search spaces, which is required for unrectified stereo matching and optical flow extraction. We also compare our results with traditional cost volume interpolation formulae as well as with state-of-the-art cost-based refinement methods, and show that the proposed formulae bring a small improvement over the state-of-the-art cost-based methods in the case of one dimensional search spaces, and a significant improvement when the search space is two dimensional.

Lian Xu, Wanli Ouyang, Mohammed Bennamoun et al.

Semantic segmentation is a challenging task in the absence of densely labelled data. Only relying on class activation maps (CAM) with image-level labels provides deficient segmentation supervision. Prior works thus consider pre-trained models to produce coarse saliency maps to guide the generation of pseudo segmentation labels. However, the commonly used off-line heuristic generation process cannot fully exploit the benefits of these coarse saliency maps. Motivated by the significant inter-task correlation, we propose a novel weakly supervised multi-task framework termed as AuxSegNet, to leverage saliency detection and multi-label image classification as auxiliary tasks to improve the primary task of semantic segmentation using only image-level ground-truth labels. Inspired by their similar structured semantics, we also propose to learn a cross-task global pixel-level affinity map from the saliency and segmentation representations. The learned cross-task affinity can be used to refine saliency predictions and propagate CAM maps to provide improved pseudo labels for both tasks. The mutual boost between pseudo label updating and cross-task affinity learning enables iterative improvements on segmentation performance. Extensive experiments demonstrate the effectiveness of the proposed auxiliary learning network structure and the cross-task affinity learning method. The proposed approach achieves state-of-the-art weakly supervised segmentation performance on the challenging PASCAL VOC 2012 and MS COCO benchmarks.

Laurent Valentin Jospin, Wray Buntine, Farid Boussaid et al.

Modern deep learning methods constitute incredibly powerful tools to tackle a myriad of challenging problems. However, since deep learning methods operate as black boxes, the uncertainty associated with their predictions is often challenging to quantify. Bayesian statistics offer a formalism to understand and quantify the uncertainty associated with deep neural network predictions. This tutorial provides an overview of the relevant literature and a complete toolset to design, implement, train, use and evaluate Bayesian Neural Networks, i.e. Stochastic Artificial Neural Networks trained using Bayesian methods.

Hamid Laga, Laurent Valentin Jospin, Farid Boussaid et al.

Estimating depth from RGB images is a long-standing ill-posed problem, which has been explored for decades by the computer vision, graphics, and machine learning communities. Among the existing techniques, stereo matching remains one of the most widely used in the literature due to its strong connection to the human binocular system. Traditionally, stereo-based depth estimation has been addressed through matching hand-crafted features across multiple images. Despite the extensive amount of research, these traditional techniques still suffer in the presence of highly textured areas, large uniform regions, and occlusions. Motivated by their growing success in solving various 2D and 3D vision problems, deep learning for stereo-based depth estimation has attracted growing interest from the community, with more than 150 papers published in this area between 2014 and 2019. This new generation of methods has demonstrated a significant leap in performance, enabling applications such as autonomous driving and augmented reality. In this article, we provide a comprehensive survey of this new and continuously growing field of research, summarize the most commonly used pipelines, and discuss their benefits and limitations. In retrospect of what has been achieved so far, we also conjecture what the future may hold for deep learning-based stereo for depth estimation research.

Ammar Mahmood, Ana Giraldo Ospina, Mohammed Bennamoun et al.

Across the globe, remote image data is rapidly being collected for the assessment of benthic communities from shallow to extremely deep waters on continental slopes to the abyssal seas. Exploiting this data is presently limited by the time it takes for experts to identify organisms found in these images. With this limitation in mind, a large effort has been made globally to introduce automation and machine learning algorithms to accelerate both classification and assessment of marine benthic biota. One major issue lies with organisms that move with swell and currents, like kelps. This paper presents an automatic hierarchical classification method (local binary classification as opposed to the conventional flat classification) to classify kelps in images collected by autonomous underwater vehicles. The proposed kelp classification approach exploits learned feature representations extracted from deep residual networks. We show that these generic features outperform the traditional off-the-shelf CNN features and the conventional hand-crafted features. Experiments also demonstrate that the hierarchical classification method outperforms the traditional parallel multi-class classifications by a significant margin (90.0% vs 57.6% and 77.2% vs 59.0%) on Benthoz15 and Rottnest datasets respectively. Furthermore, we compare different hierarchical classification approaches and experimentally show that the sibling hierarchical training approach outperforms the inclusive hierarchical approach by a significant margin. We also report an application of our proposed method to study the change in kelp cover over time for annually repeated AUV surveys.

Senjian An, Farid Boussaid, Mohammed Bennamoun et al.

By introducing sign constraints on the weights, this paper proposes sign constrained rectifier networks (SCRNs), whose training can be solved efficiently by the well known majorization-minimization (MM) algorithms. We prove that the proposed two-hidden-layer SCRNs, which exhibit negative weights in the second hidden layer and negative weights in the output layer, are capable of separating any two (or more) disjoint pattern sets. Furthermore, the proposed two-hidden-layer SCRNs can decompose the patterns of each class into several clusters so that each cluster is convexly separable from all the patterns from the other classes. This provides a means to learn the pattern structures and analyse the discriminant factors between different classes of patterns.

Senjian An, Mohammed Bennamoun, Farid Boussaid

This paper provides a theoretical justification of the superior classification performance of deep rectifier networks over shallow rectifier networks from the geometrical perspective of piecewise linear (PWL) classifier boundaries. We show that, for a given threshold on the approximation error, the required number of boundary facets to approximate a general smooth boundary grows exponentially with the dimension of the data, and thus the number of boundary facets, referred to as boundary resolution, of a PWL classifier is an important quality measure that can be used to estimate a lower bound on the classification errors. However, learning naively an exponentially large number of boundary facets requires the determination of an exponentially large number of parameters and also requires an exponentially large number of training patterns. To overcome this issue of "curse of dimensionality", compressive representations of high resolution classifier boundaries are required. To show the superior compressive power of deep rectifier networks over shallow rectifier networks, we prove that the maximum boundary resolution of a single hidden layer rectifier network classifier grows exponentially with the number of units when this number is smaller than the dimension of the patterns. When the number of units is larger than the dimension of the patterns, the growth rate is reduced to a polynomial order. Consequently, the capacity of generating a high resolution boundary will increase if the same large number of units are arranged in multiple layers instead of a single hidden layer. Taking high dimensional spherical boundaries as examples, we show how deep rectifier networks can utilize geometric symmetries to approximate a boundary with the same accuracy but with a significantly fewer number of parameters than single hidden layer nets.

Senjian An, Farid Boussaid, Mohammed Bennamoun et al.

In this paper, we introduce transformations of deep rectifier networks, enabling the conversion of deep rectifier networks into shallow rectifier networks. We subsequently prove that any rectifier net of any depth can be represented by a maximum of a number of functions that can be realized by a shallow network with a single hidden layer. The transformations of both deep rectifier nets and deep residual nets are conducted to demonstrate the advantages of the residual nets over the conventional neural nets and the advantages of the deep neural nets over the shallow neural nets. In summary, for two rectifier nets with different depths but with same total number of hidden units, the corresponding single hidden layer representation of the deeper net is much more complex than the corresponding single hidden representation of the shallower net. Similarly, for a residual net and a conventional rectifier net with the same structure except for the skip connections in the residual net, the corresponding single hidden layer representation of the residual net is much more complex than the corresponding single hidden layer representation of the conventional net.

Qiuhong Ke, Mohammed Bennamoun, Senjian An et al.

This paper presents a new method for 3D action recognition with skeleton sequences (i.e., 3D trajectories of human skeleton joints). The proposed method first transforms each skeleton sequence into three clips each consisting of several frames for spatial temporal feature learning using deep neural networks. Each clip is generated from one channel of the cylindrical coordinates of the skeleton sequence. Each frame of the generated clips represents the temporal information of the entire skeleton sequence, and incorporates one particular spatial relationship between the joints. The entire clips include multiple frames with different spatial relationships, which provide useful spatial structural information of the human skeleton. We propose to use deep convolutional neural networks to learn long-term temporal information of the skeleton sequence from the frames of the generated clips, and then use a Multi-Task Learning Network (MTLN) to jointly process all frames of the generated clips in parallel to incorporate spatial structural information for action recognition. Experimental results clearly show the effectiveness of the proposed new representation and feature learning method for 3D action recognition.