Learning to Watch: Active Video Anomaly Understanding via Interleaved Policy Optimization

Video anomaly understanding (VAU) relies on sparse, context-dependent cues. However, existing passive paradigms suffer from observational aliasing, where static sampling fails to disambiguate semantically distinct events. To overcome this, we propose $Anom\text{-}π$, a closed-loop framework that reconceptualizes video understanding as an active sequential decision-making process within a dynamic environment. Inspired by human video-reviewing behavior, this framework unifies internal cognitive reasoning and strategic evidence acquisition into an interleaved policy, utilizing temporal atomic operators such as local backtracking, temporal expansion, and fine-grained sampling to endow the model with perceptual proactivity. To learn such complex interaction strategies under video-level weak supervision, we design Interactive Direct Preference Optimization (iDPO) to achieve trajectory-level policy alignment, guided by an Active Evidence Inquiry (AEI) utility that balances task success, informative evidence acquisition, and interaction cost. This approach enables the agent to learn to actively disambiguate hypotheses while suppressing redundant exploration. Extensive experiments demonstrate that our framework, with only 2B parameters, achieves highly competitive performance, significantly outperforming state-of-the-art large-scale VAU models in complex scenarios.

Linguistic Relative Policy Optimization for Video Anomaly Reasoning

Video anomaly detection (VAD) with multimodal large language models has shown strong potential, yet most existing methods still depend on large-scale annotations or expert-designed priors, limiting their ability to acquire anomaly knowledge with as little human intervention as possible. To address this, we propose Linguistic Relative Policy Optimization (LRPO), which distills group-relative semantic advantages from multiple reasoning trajectories into a linguistically expressed anomaly experience prior, and adapts the model by injecting this prior into the context to steer its output distribution without any parameter updates. LRPO builds two complementary experience representations: general experience captures transferable anomaly preferences across scenarios, while scenario experience models context-dependent anomaly rules for targeted refinement. To further improve the learned experience, we introduce an anomaly alignment reward that guides trajectory optimization to match human risk preferences and reinforce temporally grounded reasoning. Extensive experiments on XD-Violence, UCF-Crime, and UBnormal demonstrate that LRPO significantly outperforms existing state-of-the-art methods under tuning-free settings.

A2Seek: Towards Reasoning-Centric Benchmark for Aerial Anomaly Understanding

While unmanned aerial vehicles (UAVs) offer wide-area, high-altitude coverage for anomaly detection, they face challenges such as dynamic viewpoints, scale variations, and complex scenes. Existing datasets and methods, mainly designed for fixed ground-level views, struggle to adapt to these conditions, leading to significant performance drops in drone-view scenarios. To bridge this gap, we introduce A2Seek (Aerial Anomaly Seek), a large-scale, reasoning-centric benchmark dataset for aerial anomaly understanding. This dataset covers various scenarios and environmental conditions, providing high-resolution real-world aerial videos with detailed annotations, including anomaly categories, frame-level timestamps, region-level bounding boxes, and natural language explanations for causal reasoning. Building on this dataset, we propose A2Seek-R1, a novel reasoning framework that generalizes R1-style strategies to aerial anomaly understanding, enabling a deeper understanding of "Where" anomalies occur and "Why" they happen in aerial frames. To this end, A2Seek-R1 first employs a graph-of-thought (GoT)-guided supervised fine-tuning approach to activate the model's latent reasoning capabilities on A2Seek. Then, we introduce Aerial Group Relative Policy Optimization (A-GRPO) to design rule-based reward functions tailored to aerial scenarios. Furthermore, we propose a novel "seeking" mechanism that simulates UAV flight behavior by directing the model's attention to informative regions. Extensive experiments demonstrate that A2Seek-R1 achieves up to a 22.04% improvement in AP for prediction accuracy and a 13.9% gain in mIoU for anomaly localization, exhibiting strong generalization across complex environments and out-of-distribution scenarios. Our dataset and code will be released at https://hayneyday.github.io/A2Seek/.

PiercingEye: Dual-Space Video Violence Detection with Hyperbolic Vision-Language Guidance

Existing weakly supervised video violence detection (VVD) methods primarily rely on Euclidean representation learning, which often struggles to distinguish visually similar yet semantically distinct events due to limited hierarchical modeling and insufficient ambiguous training samples. To address this challenge, we propose PiercingEye, a novel dual-space learning framework that synergizes Euclidean and hyperbolic geometries to enhance discriminative feature representation. Specifically, PiercingEye introduces a layer-sensitive hyperbolic aggregation strategy with hyperbolic Dirichlet energy constraints to progressively model event hierarchies, and a cross-space attention mechanism to facilitate complementary feature interactions between Euclidean and hyperbolic spaces. Furthermore, to mitigate the scarcity of ambiguous samples, we leverage large language models to generate logic-guided ambiguous event descriptions, enabling explicit supervision through a hyperbolic vision-language contrastive loss that prioritizes high-confusion samples via dynamic similarity-aware weighting. Extensive experiments on XD-Violence and UCF-Crime benchmarks demonstrate that PiercingEye achieves state-of-the-art performance, with particularly strong results on a newly curated ambiguous event subset, validating its superior capability in fine-grained violence detection.

EHGCN: Hierarchical Euclidean-Hyperbolic Fusion via Motion-Aware GCN for Hybrid Event Stream Perception

Event cameras, with microsecond temporal resolution and high dynamic range (HDR) characteristics, emit high-speed event stream for perception tasks. Despite the recent advancement in GNN-based perception methods, they are prone to use straightforward pairwise connectivity mechanisms in the pure Euclidean space where they struggle to capture long-range dependencies and fail to effectively characterize the inherent hierarchical structures of non-uniformly distributed event stream. To this end, in this paper we propose a novel approach named EHGCN, which is a pioneer to perceive event stream in both Euclidean and hyperbolic spaces for event vision. In EHGCN, we introduce an adaptive sampling strategy to dynamically regulate sampling rates, retaining discriminative events while attenuating chaotic noise. Then we present a Markov Vector Field (MVF)-driven motion-aware hyperedge generation method based on motion state transition probabilities, thereby eliminating cross-target spurious associations and providing critically topological priors while capturing long-range dependencies between events. Finally, we propose a Euclidean-Hyperbolic GCN to fuse the information locally aggregated and globally hierarchically modeled in Euclidean and hyperbolic spaces, respectively, to achieve hybrid event perception. Experimental results on event perception tasks such as object detection and recognition validate the effectiveness of our approach.

Shape-centered Representation Learning for Visible-Infrared Person Re-identification

Current Visible-Infrared Person Re-Identification (VI-ReID) methods prioritize extracting distinguishing appearance features, ignoring the natural resistance of body shape against modality changes. Initially, we gauged the discriminative potential of shapes by a straightforward concatenation of shape and appearance features. However, two unresolved issues persist in the utilization of shape features. One pertains to the dependence on auxiliary models for shape feature extraction in the inference phase, along with the errors in generated infrared shapes due to the intrinsic modality disparity. The other issue involves the inadequately explored correlation between shape and appearance features. To tackle the aforementioned challenges, we propose the Shape-centered Representation Learning framework (ScRL), which focuses on learning shape features and appearance features associated with shapes. Specifically, we devise the Shape Feature Propagation (SFP), facilitating direct extraction of shape features from original images with minimal complexity costs during inference. To restitute inaccuracies in infrared body shapes at the feature level, we present the Infrared Shape Restitution (ISR). Furthermore, to acquire appearance features related to shape, we design the Appearance Feature Enhancement (AFE), which accentuates identity-related features while suppressing identity-unrelated features guided by shape features. Extensive experiments are conducted to validate the effectiveness of the proposed ScRL. Achieving remarkable results, the Rank-1 (mAP) accuracy attains 76.1%, 71.2%, 92.4% (72.6%, 52.9%, 86.7%) on the SYSU-MM01, HITSZ-VCM, RegDB datasets respectively, outperforming existing state-of-the-art methods.