WACV 2026 Papers

With the rapid deployment of vision systems on edge devices, energy-efficient and temporally aware image deraining models are increasingly needed. We propose SpikeRain, a spiking neural network (SNN) that achieves competitive deraining performance with substantially lower computational cost than conventional artificial neural networks (ANNs). Unlike ANN-based approaches with dense activations and high memory demands, SpikeRain leverages the event-driven sparse-firing nature of spiking neurons for efficient temporal integration and contextual learning. Built on an encoder-decoder framework, SpikeRain incorporates three spiking native modules: a Dense Spiking Residual Block (DSRB) for temporal integration and feature reuse, a Multi-Dimensional Spiking Attention (MDSA) module to model temporal channel spatial dependencies, and an Adaptive Residual Feature Enhancement (ARFE) block with gated attention to refine salient features. Experiments on synthetic and real-world benchmarks show that SpikeRain achieves state-of-the-art PSNR and SSIM while reducing parameters by approximately 40% and FLOPs by approximately 89%, with energy efficiency on par with existing SNN methods. These results highlight the potential of SNNs for real-time low-power image restoration on neuromorphic platforms. SpikeRain code is available on \href https://github.com/tanvirnwu/SpikeRain GitHub .

Multimodal Large Language Models (MLLMs) have achieved impressive performance across a variety of vision-language tasks. However, their internal working mechanisms remain largely underexplored. In his work, we introduce FG-TRACER, a framework designed to analyze the information flow between visual and textual modalities in MLLMs in free-form generation. Notably, our numerically stabilized computational method enables the first systematic analysis of multimodal information flow in underexplored domains such as image captioning and chain-of-thought (CoT) reasoning. We apply FG-TRACER to three state-of-the-art MLLMs--LLaVA 1.5, LLaMA 3.2-Vision, and Qwen 2.5-VL--across three vision-language benchmarks--TextVQA, COCO 2014, and ChartQA--and we conduct a word-level analysis of multimodal integration. Our findings uncover distinct patterns of multimodal fusion across models and tasks, demonstrating that fusion dynamics are both model- and task-dependent. Overall, FG-TRACER offers a robust methodology for probing the internal mechanisms of MLLMs in free-form settings, providing new insights into their multimodal reasoning strategies. Our source code is publicly available at https://github.com/AImageLab-zip/FG-TRACER

Accurate identification of individual cows is essential to precision dairy farming. While computer vision offers a non-invasive alternative to ear tags and RFID systems, its practical deployment remains limited by the need for zero-shot identification in dynamic herds where test identities are unseen during training. In this work, we propose OpenCowID, a unified framework that addresses this challenge. First, we introduce a stochastic cow coat synthesis pipeline that efficiently generates large-scale, diverse images. Second, using the generated large-scale high-quality data, we present a centroid-guided feature learning strategy that forms a well-structured embedding space using virtual class centroids, enabling generalization to unseen identities. OpenCowID achieves state-of-the-art zero-shot and open-set identification on real-world cow benchmarks, without requiring any real labeled training data. This work contributes to the advancement of automated livestock monitoring, enabling robust, non-invasive identification. Code is available at https://github.com/neis-lab/OpenCowID.

Scoliosis presents significant diagnostic challenges, especially in its early stages due to structural deformities of the spine. The most common type, Adolescent Idiopathic Scoliosis (AIS), typically appears during rapid growth periods between ages 10 and 15, accounting for about 80-85% of all scoliosis cases. Currently, diagnosis mainly involves repeated X-rays and clinical evaluations, which are costly and expose patients to frequent radiation. To address these challenges, we propose ScoliGaitX, a novel non-invasive system for assessing scoliosis through gait video analysis. Our approach uses a multi-stream deep learning model trained specifically for scoliosis classification using the Scoliosis1K dataset. We integrate three different gait modalities--(i) silhouettes to capture appearance, (ii) optical flow for motion analysis, and (iii) GEI-subtracted sequences to highlight deviations from normal gait--to effectively identify gait abnormalities associated with scoliosis. A key component of our proposal is the Align Gate Fuse (AGF) module, designed to efficiently learn relationships between different modalities. It achieves this by adaptively assigning importance to each modality through a lightweight global gating mechanism. Our experimental results demonstrate that ScoliGaitX significantly outperforms existing methods, achieving an accuracy of 89.05%, which is 7.05% higher than the previous best approach (ScolNet-MT), while maintaining excellent specificity. This highlights the promise of our method for providing early, radiation-free scoliosis assessment.

This paper proposes a novel metric for Human Motion Prediction (HMP). Since a single past sequence can lead to multiple possible futures, a probabilistic HMP method predicts such multiple motions. While a single motion predicted by a deterministic method is evaluated only with the difference from its ground truth motion, multiple predicted motions should also be evaluated based on their distribution. For this evaluation, this paper focuses on the following two criteria. (a) Coverage: motions should be distributed among multiple motion modes to cover diverse possibilities. (b) Validity: motions should be kinematically valid as future motions observable from a given past motion. However, existing metrics simply appreciate widely distributed motions even if these motions are observed in a single mode and kinematically invalid. To resolve these disadvantages, this paper proposes a Multimodality-aware Metric using Clustering-based Modes (MMCM). For (a) coverage, MMCM divides a motion space into several clusters, each of which is regarded as a mode. These modes are used to explicitly evaluate whether predicted motions are distributed among multiple modes. For (b) validity, MMCM identifies valid modes by collecting possible future motions from a motion dataset. Our experiments validate that our clustering yields sensible mode definitions and that MMCM accurately scores multimodal predictions. Code: https://github.com/placerkyo/MMCM

Rapid progress in large vision-language models(LVLMs) has achieved unprecedented performance in vision-language tasks. However, due to the strong prior of large language models (LLMs) and misaligned attention across modalities, LVLMs often generate outputs inconsistent with visual content - termed hallucination. To address this, we propose Scalpel, a method that reduces hallucination by refining attention activation distributions toward more credible regions. Scalpel predicts trusted attention directions for each head in Transformer layers during inference and adjusts activations accordingly. It employs a Gaussian mixture model to capture multi-peak distributions of attention in trust and hallucination manifolds, and uses entropic optimal transport (equivalent to Schrodinger bridge problem) to map Gaussian components precisely. During mitigation, Scalpel dynamically adjusts intervention strength and direction based on component membership and mapping relationships between hallucination and trust activations. Extensive experiments across multiple datasets and benchmarks demonstrate that Scalpel effectively mitigates hallucinations, outperforming previous methods and achieving state-of-the-art performance. Moreover, Scalpel is model- and data-agnostic, requiring no additional computation, only a single decoding step.

Parameter-efficient fine-tuning (PEFT) significantly reduces computational and memory costs by updating only a small subset of the model's parameters, enabling faster adaptation to new tasks with minimal loss in performance. Previous studies have introduced PEFTs tailored for point cloud data, as general approaches are suboptimal. To further reduce the number of trainable parameters, we propose a point-cloud-specific PEFT, termed Graph Features Tuning (GFT), which learns a dynamic graph from initial tokenized inputs of the transformer using a lightweight graph convolution network and passes these graph features to deeper layers via skip connections and efficient cross-attention modules. Extensive experiments on object classification and segmentation tasks show that GFT operates in the same domain, rivalling existing methods, while reducing the trainable parameters. Code is available at https://github.com/manishdhakal/GFT.

Custom Diffusion Models (CDMs) have gained significant attention due to their remarkable ability to personalize generative processes. However, existing CDMs suffer from catastrophic forgetting when continuously learning new concepts. Most prior works attempt to mitigate this issue under the sequential learning setting with a fixed order of concept inflow and neglect inter-concept interactions. In this work, we propose a novel framework - Forget Less by Learning Together (FL2T) - that enables concurrent and order-agnostic concept learning while addressing catastrophic forgetting. Specifically, we introduce a set-invariant inter-concept learning module where proxies guide feature selection across concepts, facilitating improved knowledge retention and transfer. By leveraging inter-concept guidance, our approach preserves old concepts while efficiently incorporating new ones. Extensive experiments, across three datasets, demonstrates that our method significantly improves concept retention and mitigates catastrophic forgetting, highlighting the effectiveness of inter-concept catalytic behavior in incremental concept learning of ten tasks with atleast 2% gain on average Image Alignment scores.

Interpretability is essential in Whole Slide Image (WSI) analysis for computational pathology, where understanding model predictions helps build trust in AI-assisted diagnostics. While Integrated Gradients (IG) and related attribution methods have shown promise, applying them directly to WSIs introduces challenges due to their high-resolution nature. These methods capture model decision patterns but may overlook class-discriminative signals that are crucial for distinguishing between tumor subtypes. In this work, we introduce Contrastive Integrated Gradients (CIG), a novel attribution method that enhances interpretability by computing contrastive gradients in logit space. First, CIG highlights class-discriminative regions by comparing feature importance relative to a reference class, offering sharper differentiation between tumor and non-tumor areas. Second, CIG satisfies the axioms of integrated attribution, ensuring consistency and theoretical soundness. Third, we propose two attribution quality metrics, MIL-AIC and MIL-SIC, which measure how predictive information and model confidence evolve with access to salient regions, particularly under weak supervision. We validate CIG across three datasets spanning distinct cancer types: CAMELYON16 (breast cancer metastasis in lymph nodes), TCGA-RCC (renal cell carcinoma), and TCGA-Lung (lung cancer). Experimental results demonstrate that CIG yields more informative attributions both quantitatively, using MIL-AIC and MIL-SIC, and qualitatively, through visualizations that align closely with ground truth tumor regions, underscoring its potential for interpretable and trustworthy WSI-based diagnostics.

Existing text-to-motion (T2M) generation methods primarily rely on regression-based objectives, such as minimizing positional errors. However, they lack effective semantic supervision and correction mechanisms, often leading to substantial misalignment between text and motion. To address this, we propose Aligned Text-to-Motion (ATM), a semantics-aware generation framework that automatically identifies and corrects text-motion misalignment. ATM incorporates two key components: (1) Inter-motion alignment, which detects semantic contradictions across motions and applies adaptive corrections based on the degree of semantic discrepancy, flexibly handing diverse misalignments and ensuring global text-motion consistency; (2) Intra-motion alignment, which refines locally missing or inaccurate motion semantics in an unsupervised manner by inferring semantic proxies, effectively addressing the absence of localized textual annotations. ATM is model-agnostic and can be seamlessly integrated into various T2M methods as a plug-and-play module. Extensive experiments on HumanML3D and KIT demonstrate that ATM consistently improves both generation quality and text-motion alignment. Code is available at https://github.com/ke-han-aca/ATM.git.