Is Video Anomaly Detection Misframed? Evidence from LLM-Based and Multi-Scene Models
For researchers in video anomaly detection, this paper highlights a critical misalignment between current trends and real-world requirements, but the argument is largely conceptual and lacks novel empirical results.
The paper argues that current video anomaly detection (VAD) research is misframed by focusing on multi-scene generalization and large language models, which shift the task toward semantic action recognition rather than detecting deviations from scene-specific normality. Through analyses and evaluations, they show this leads to poor spatial localization and semantic bias, advocating for a return to single-scene, spatially-aware, and explainable approaches.
Recent video anomaly detection research has expanded rapidly with an emphasis on general models of normality intended to work across many different scenes. While this focus has led to improvements in scalability and multi-scene generalization, it has also shifted the field away from modeling the scene-specific and context-dependent nature of normal behavior. Contemporary approaches frequently rely on video-level weak supervision and opaque pretrained representations from multi-modal large language models (MLLMs), which encourage models to respond to familiar semantic anomaly categories rather than to deviations from the normal patterns of a particular environment. This trend suppresses spatial localization, introduces semantic bias, and reduces anomaly detection to a form of action recognition. In this paper, we examine whether these prevailing formulations align with the core requirements of real-world VAD, which is typically performed within a single scene where normality is determined by local geometry, semantics, and activity patterns. Through targeted visual analyses and empirical evaluations, we demonstrate the practical consequences of these limitations and show that meaningful progress in VAD requires renewed focus on single-scene, spatially-aware, and explainable formulations that capture the nuanced structure of normality within individual environments.