NeurIPS 2025 Papers

This paper presents a novel problem, interactive anomaly detection (AD) for articulated objects, and introduces a tailored solution that detects functional anomalies by integrating vision, interaction, and anticipation. Unlike traditional AD methods that rely on passive visual observations, our approach actively manipulates objects to reveal anomalies that would otherwise remain hidden. Our method learns to generate a sequence of actions to interact exclusively with normal objects and to anticipate the resulting normal motion. During inference, the model applies predicted actions to the object and compares the observed motion with the anticipated motion to detect anomalies. Additionally, we introduce a new benchmark, PartNet-IAD, for interactive AD, which includes articulated objects with realistic functional anomalies. Experiments show strong generalization to detect anomalies in both seen and unseen object categories. Code and dataset will be released.

Interpreting deep neural networks through concept-based explanations offers a bridge between low-level features and high-level human-understandable semantics. However, existing automatic concept discovery methods often fail to align these extracted concepts with the model’s true decision-making process, thereby compromising explanation faithfulness. In this work, we propose FACE (Faithful Automatic Concept Extraction), a novel framework that combines Non-negative Matrix Factorization (NMF) with a Kullback-Leibler (KL) divergence regularization term to ensure alignment between the model’s original and concept-based predictions. Unlike prior methods that operate solely on encoder activations, FACE incorporates classifier supervision during concept learning, enforcing predictive consistency and enabling faithful explanations. We provide theoretical guarantees showing that minimizing the KL divergence bounds the deviation in predictive distributions, thereby promoting faithful local linearity in the learned concept space. Systematic evaluations on ImageNet, COCO, and CelebA datasets demonstrate that FACE outperforms existing methods across faithfulness and sparsity metrics.

In the era of artificial intelligence (AI), pedestrian privacy is increasingly at risk. In research areas such as autonomous driving, computer vision, and surveillance, large datasets are often collected in public spaces, capturing pedestrians without consent or anonymization. These datasets are used to train systems that can identify, track, and analyze individuals, often without their knowledge. Although various technical methods and regional regulations have been proposed to address this issue, existing solutions are either insufficient to protect privacy or compromise data utility, thereby limiting their effectiveness for research. In this paper, we argue that more effort is needed to protect pedestrian privacy in the era of AI while maintaining data utility. We call on the AI and computer vision communities to take pedestrian privacy seriously and to rethink how pedestrian data are collected and anonymized. Collaboration with experts in law and ethics will also be essential for the responsible development of AI. Without stronger action, it will become increasingly difficult for individuals to protect their privacy, and public trust in AI may decline.

We present Eve3D, a novel framework for dense 3D reconstruction based on 3D Gaussian Splatting (3DGS). While most existing methods rely on imperfect priors derived from pre-trained vision models, Eve3D fully leverages these priors by jointly optimizing both them and the 3DGS backbone. This joint optimization creates a mutually reinforcing cycle: the priors enhance the quality of 3DGS, which in turn refines the priors, further improving the reconstruction. Additionally, Eve3D introduces a novel optimization step based on bundle adjustment, overcoming the limitations of the highly local supervision in standard 3DGS pipelines. Eve3D achieves state-of-the-art results in surface reconstruction and novel view synthesis on the Tanks & Temples, DTU, and Mip-NeRF360 datasets. while retaining fast convergence, highlighting an unprecedented trade-off between accuracy and speed.

Novel view synthesis is an important problem with many applications, including AR/VR, gaming, and robotic simulations. With the recent rapid development of Neural Radiance Fields (NeRFs) and 3D Gaussian Splatting (3DGS) methods, it is becoming difficult to keep track of the current state of the art (SoTA) due to methods using different evaluation protocols, codebases being difficult to install and use, and methods not generalizing well to novel 3D scenes. In our experiments, we show that even tiny differences in the evaluation protocols of various methods can artificially boost the performance of these methods. This raises questions about the validity of quantitative comparisons performed in the literature. To address these questions, we propose NerfBaselines, an evaluation framework which provides consistent benchmarking tools, ensures reproducibility, and simplifies the installation and use of various methods. We validate our implementation experimentally by reproducing the numbers reported in the original papers. For improved accessibility, we release a web platform that compares commonly used methods on standard benchmarks. We strongly believe NerfBaselines is a valuable contribution to the community as it ensures that quantitative results are comparable and thus truly measure progress in the field of novel view synthesis.

Transformers have driven remarkable breakthroughs in natural language processing and computer vision, yet their standard attention mechanism still imposes $O(N^2)$ complexity, hindering scalability to longer sequences. We introduce Circular-convolutional ATtention (CAT), a Fourier-based approach that efficiently applies circular convolutions to reduce complexity without sacrificing representational power. CAT achieves $O(N \log N)$ computations, requires fewer learnable parameters by streamlining fully connected layers, and introduces no heavier operations, resulting in consistent accuracy improvements and about a 10\% speedup in naive PyTorch implementations. Based on the engineering-isomorphic transformer framework, CAT's design not only offers practical efficiency and ease of implementation, but also provides insights to guide the development of future high-performance Transformer architectures. Finally, our ablation studies highlight the key conditions underlying CAT’s success, shedding light on broader principles for scalable attention mechanisms.

What is a gait? Appearance-based gait networks consider a gait as the human shape and motion information from images. Model-based gait networks treat a gait as the human inherent structure from points. However, the considerations remain vague for humans to comprehend truly. In this work, we introduce a novel paradigm Vocabulary-Guided Gait Recognition, dubbed Gait-World, which attempts to explore gait concepts through human vocabularies with Vision-Language Models (VLMs). Despite VLMs have achieved the remarkable progress in various vision tasks, the cognitive capability regarding gait modalities remains limited. The success element in Gait-World is the proper vocabulary prompt where this paradigm carefully selects gait cycle actions as Vocabulary Base, bridging the gait and vocabulary feature spaces and further promoting human understanding for the gait. How to extract gait features? Although previous gait networks have made significant progress, learning solely from gait modalities on limited gait databases makes it difficult to learn robust gait features for practicality. Therefore, we propose the first Gait-World model, dubbed $\alpha$-Gait, which guides the gait network learning with universal vocabulary knowledge from VLMs. However, due to the heterogeneity of the modalities, directly integrating vocabulary and gait features is highly challenging as they reside in different embedding spaces. To address the issues, $\alpha$-Gait designs Vocabulary Relation Mapper and Gait Fine-grained Detector to map and establish vocabulary relations in the gait space for detecting corresponding gait features. Extensive experiments on CASIA-B, CCPG, SUSTech1K, Gait3D and GREW reveal the potential value and research directions of vocabulary information from VLMs in the gait field.

Accurate 3D reconstruction of hands and instruments is critical for vision-based analysis of ophthalmic microsurgery, yet progress has been hampered by the lack of realistic, large-scale datasets and reliable annotation tools. In this work, we introduce OphNet-3D, the first extensive RGB-D dynamic 3D reconstruction dataset for ophthalmic surgery, comprising 41 sequences from 40 surgeons and totaling 7.1 million frames, with fine-grained annotations of 12 surgical phases, 10 instrument categories, dense MANO hand meshes, and full 6-DoF instrument poses. To scalably produce high-fidelity labels, we design a multi-stage automatic annotation pipeline that integrates multi-view data observation, data-driven motion prior with cross-view geometric consistency and biomechanical constraints, along with a combination of collision-aware interaction constraints for instrument interactions. Building upon OphNet-3D, we establish two challenging benchmarks—bimanual hand pose estimation and hand–instrument interaction reconstruction—and propose two dedicated architectures: H-Net for dual-hand mesh recovery and OH-Net for joint reconstruction of two-hand–two-instrument interactions. These models leverage a novel spatial reasoning module with weak-perspective camera modeling and collision-aware center-based representation. Both architectures outperform existing methods by substantial margins, achieving improvements of over 2mm in Mean Per Joint Position Error (MPJPE) and up to 23\% in ADD-S metrics for hand and instrument reconstruction, respectively.

Spiking Neural Networks promise brain-inspired and energy-efficient computation by transmitting information through binary (0/1) spikes. Yet, their performance still lags behind that of artificial neural networks, often assumed to result from information loss caused by sparse and binary activations. In this work, we challenge this long-standing assumption and reveal a previously overlooked frequency bias: **spiking neurons inherently suppress high-frequency components and preferentially propagate low-frequency information.** This frequency-domain imbalance, we argue, is the root cause of degraded feature representation in SNNs. Empirically, on Spiking Transformers, adopting Avg-Pooling (low-pass) for token mixing lowers performance to 76.73% on Cifar-100, whereas replacing it with Max-Pool (high-pass) pushes the top-1 accuracy to 79.12%. Accordingly, we introduce **Max-Former** that restores high-frequency signals through two frequency-enhancing operators: (1) extra Max-Pool in patch embedding, and (2) Depth-Wise Convolution in place of self-attention. Notably, **Max-Former** attains 82.39% top-1 accuracy on ImageNet using only 63.99M parameters, surpassing Spikformer (74.81%, 66.34M) by +7.58%. Extending our insight beyond transformers, our **Max-ResNet-18** achieves state-of-the-art performance on convolution-based benchmarks: 97.17% on CIFAR-10 and 83.06% on CIFAR-100. We hope this simple yet effective solution inspires future research to explore the distinctive nature of spiking neural networks. Code is available: https://github.com/bic-L/MaxFormer.

Video Temporal Grounding (VTG) aims to localize relevant segments in untrimmed videos based on natural language queries and has seen notable progress in recent years. However, most existing methods suffer from two critical limitations. First, they are prone to learning superficial co-occurrence patterns—such as associating specific objects or phrases with certain events—induced by dataset biases, which ultimately degrades their semantic understanding abilities. Second, they typically assume that relevant segments always exist in the video, an assumption misaligned with real-world scenarios where queried content may be absent. Fortunately, causal inference offers a natural solution to the above-mentioned issues by disentangling dataset-induced biases and enabling counterfactual reasoning about query relevance. To this end, we propose CausalVTG, a novel framework that explicitly integrates causal reasoning into VTG. Specifically, we introduce a causality-aware disentangled encoder (CADE) based on front-door adjustment to mitigate confounding biases in visual and textual modalities. To better capture temporal granularity, we design a multi-scale temporal perception module (MSTP) that reconstructs query-conditioned video features at multiple resolutions. Additionally, a counterfactual contrastive learning objective is employed to help the model discern whether a query is truly grounded in a video. Extensive experiments on five widely-used benchmarks demonstrate that CausalVTG outperforms state-of-the-art methods, achieving higher localization precision under stricter IoU thresholds and more accurately identifying whether a query is truly grounded in the video. These results demonstrate both the effectiveness and generalizability of proposed CausalVTG.

Deep Neural Networks (DNNs) have demonstrated remarkable performance across various domains, including computer vision and natural language processing. However, they often struggle to accurately quantify their predictions' uncertainty, limiting their broader adoption in critical industrial applications. Uncertainty Quantification (UQ) for Deep Learning seeks to address this challenge by providing methodologies to improve the reliability of uncertainty estimates. While numerous techniques have been proposed, a unified tool remains lacking that offers a seamless workflow for evaluating and integrating these methods. To bridge this gap, we introduce **Torch-Uncertainty**, a *PyTorch* and *Lightning* framework designed to streamline the training and evaluation of DNNs with UQ techniques. In this paper, we outline the foundational principles of our library and present comprehensive experimental results that benchmark a diverse set of UQ methods across classification, segmentation, and regression tasks. Our library is available at: https://github.com/ENSTA-U2IS-AI/torch-uncertainty.

Long-form video understanding presents significant challenges due to extensive temporal-spatial complexity and the difficulty of question answering under such extended contexts. While Large Language Models (LLMs) have demonstrated considerable advancements in video analysis capabilities and long context handling, they continue to exhibit limitations when processing information-dense hour-long videos. To overcome such limitations, we propose the $\textbf{D}eep \ \textbf{V}ideo \ \textbf{D}iscovery \ (\textbf{DVD})$ agent to leverage an $\textit{agentic search}$ strategy over segmented video clips. Different from previous video agents manually designing a rigid workflow, our approach emphasizes the autonomous nature of agents. By providing a set of search-centric tools on multi-granular video database, our DVD agent leverages the advanced reasoning capability of LLM to plan on its current observation state, strategically selects tools to orchestrate adaptive workflow for different queries in light of the gathered information. We perform comprehensive evaluation on multiple long video understanding benchmarks that demonstrates our advantage. Our DVD agent achieves state-of-the-art performance on the challenging LVBench dataset, reaching an accuracy of $\textbf{74.2\%}$, which substantially surpasses all prior works, and further improves to $\textbf{76.0\%}$ with transcripts.

Out-of-distribution (OOD) fine-tuning has emerged as a promising approach for anomaly segmentation. Current OOD fine-tuning strategies typically employ global-level objectives, aiming to guide segmentation models to accurately predict a large number of anomaly pixels. However, these strategies often perform poorly on small anomalies. To address this issue, we propose an instance-level OOD fine-tuning framework, dubbed LNOIB (Leaving No OOD Instance Behind). We start by theoretically analyzing why global-level objectives fail to segment small anomalies. Building on this analysis, we introduce a simple yet effective instance-level objective. Moreover, we propose a feature separation objective to explicitly constrain the representations of anomalies, which are prone to be smoothed by their in-distribution (ID) surroundings. LNOIB integrates these objectives to enhance the segmentation of small anomalies and serves as a paradigm adaptable to existing OOD fine-tuning strategies, without introducing additional inference cost. Experimental results show that integrating LNOIB into various OOD fine-tuning strategies yields significant improvements, particularly in component-level results, highlighting its strength in comprehensive anomaly segmentation.

We introduce MusclePose as an end-to-end learnable physics-infused 3D human pose estimator that incorporates muscle-dynamics modeling to infer human dynamics from monocular video. Current physics pose estimators aim to predict physically plausible poses by enforcing the underlying dynamics equations that govern motion. Since this is an underconstrained problem without force-annotated data, methods often estimate kinetics with external physics optimizers that may not be compatible with existing learning frameworks, or are too slow for real-time inference. While more recent methods use a regression-based approach to overcome these issues, the estimated kinetics can be seen as auxiliary predictions, and may not be physically plausible. To this end, we build on existing regression-based approaches, and aim to improve the biofidelity of kinetic inference with a multihypothesis approach --- by inferring joint torques via Lagrange’s equations and via muscle dynamics modeling with muscle torque generators. Furthermore, MusclePose predicts detailed human anthropometrics based on values from biomechanics studies, in contrast to existing physics pose estimators that construct their human models with shape primitives. We show that MusclePose is competitive with existing 3D pose estimators in positional accuracy, while also able to infer plausible human kinetics and muscle signals consistent with values from biomechanics studies, without requiring an external physics engine.

Estimating scene lighting from a single image or video remains a longstanding challenge in computer vision and graphics. Learning-based approaches are constrained by the scarcity of ground-truth HDR environment maps, which are expensive to capture and limited in diversity. While recent generative models offer strong priors for image synthesis, lighting estimation remains difficult due to its reliance on indirect visual cues, the need to infer global (non-local) context, and the recovery of high-dynamic-range outputs. We propose LuxDiT, a novel data-driven approach that fine-tunes a video diffusion transformer to generate HDR environment maps conditioned on visual input. Trained on a large synthetic dataset with diverse lighting conditions, our model learns to infer illumination from indirect visual cues and generalizes effectively to real-world scenes. To improve semantic alignment between the input and the predicted environment map, we introduce a low-rank adaptation finetuning strategy using a collected dataset of HDR panoramas. Our method produces accurate lighting predictions with realistic angular high-frequency details, outperforming existing state-of-the-art techniques in both quantitative and qualitative evaluations.

The real world is dynamic, yet most image fusion methods process static frames independently, ignoring temporal correlations in videos and leading to flickering and temporal inconsistency. To address this, we propose Unified Video Fusion (UniVF), a novel and unified framework for video fusion that leverages multi-frame learning and optical flow-based feature warping for informative, temporally coherent video fusion. To support its development, we also introduce Video Fusion Benchmark (VF-Bench), the first comprehensive benchmark covering four video fusion tasks: multi-exposure, multi-focus, infrared-visible, and medical fusion. VF-Bench provides high-quality, well-aligned video pairs obtained through synthetic data generation and rigorous curation from existing datasets, with a unified evaluation protocol that jointly assesses the spatial quality and temporal consistency of video fusion. Extensive experiments show that UniVF achieves state-of-the-art results across all tasks on VF-Bench. Project page: [vfbench.github.io](https://vfbench.github.io).

3D Human Pose Estimation (HPE) is a fundamental task in the computer vision. Generalization in 3D HPE task is crucial due to the need for robustness across diverse environments and datasets. Existing methods often focus on learning relationships between joints to enhance the generalization capability, but the role of the loss landscape, which is closely tied to generalization, remains underexplored. In this paper, we empirically visualize the loss landscape of the 3D HPE task, revealing its complexity and the challenges it poses for optimization. To address this, we first introduce a simple adaptive scaling mechanism that smooths the loss landscape. We further observe that different solutions on this smoothed loss landscape exhibit varying generalization behaviors. Based on this insight, we propose an efficient ensemble approach that combines diverse solutions on the smooth loss landscape induced by our adaptive scaling mechanism. Extensive experimental results demonstrate that our approach improves the generalization capability of 3D HPE models, and can be easily applied, regardless of model architecture, with consistent performance gains.

This study identifies that visual-related functional decoding is distributed across different decoder layers in Multimodal Large Language Models (MLLMs). Typically, each function, such as counting, grounding, or OCR recognition, narrows down to two or three layers, which we define as Vision Function Layers (VFL). Additionally, the depth and its order of different VFLs exhibits a consistent pattern across different MLLMs, which is well-aligned with human behaviors (e.g., recognition occurs first, followed by counting, and then grounding). These findings are derived from Visual Token Swapping, our novel analytical framework that modifies targeted KV cache entries to precisely elucidate layer-specific functions during decoding. Furthermore, these insights offer substantial utility in tailoring MLLMs for real-world downstream applications. For instance, when LoRA training is selectively applied to VFLs whose functions align with the training data, VFL-LoRA not only outperform full-LoRA but also prevent out-of-domain function forgetting. Moreover, by analyzing the performance differential on training data when particular VFLs are ablated, VFL-select automatically classifies data by function, enabling highly efficient data selection to directly bolster corresponding capabilities. Consequently, VFL-select surpasses human experts in data selection, and achieves 98% of full-data performance with only 20% of the original dataset. This study delivers deeper comprehension of MLLM visual processing, fostering the creation of more efficient, interpretable, and robust models.

Recent advances in multimodal training have significantly improved the integration of image understanding and generation within a unified model. This study investigates how vision-language models (VLMs) handle image-understanding tasks, focusing on how visual information is processed and transferred to the textual domain. We compare *native multimodal VLMs*, models trained from scratch on multimodal data to generate both text and images, and *non-native multimodal VLMs*, models adapted from pre-trained large language models or capable of generating only text, highlighting key differences in information flow. We find that in native multimodal VLMs, image and text embeddings are more separated within the residual stream. Moreover, VLMs differ in how visual information reaches text: non-native multimodal VLMs exhibit a distributed communication pattern, where information is exchanged through multiple image tokens, whereas models trained natively for joint image and text generation tend to rely on a single post-image token that acts as a *narrow gate* for visual information. We show that ablating this single token significantly deteriorates image-understanding performance, whereas targeted, token-level interventions reliably steer image semantics and downstream text with fine-grained control.