Cog-VADU: A Training-Free Cognitive Reasoning Framework for Video Anomaly Detection and Understanding

2026-10-01

AI Summary

This study addresses the limitations of existing video anomaly detection methods, which typically rely on task-specific training data, while zero-shot approaches lack temporal continuity and structured reasoning capabilities. To overcome these challenges, this work proposes Cog-VADU, a training-free framework that reformulates anomaly detection as a sequential cognitive reasoning task. Specifically, it introduces the Chain-of-Anomaly-Detection-Thought Prompting (CoADTP) strategy, which maintains implicit temporal memory through recursive reasoning, and incorporates cross-modal re-ranking to enhance semantic consistency. Built upon large vision-language models, the proposed method is model-agnostic and achieves state-of-the-art zero-shot performance across multiple benchmarks. Ultimately, Cog-VADU enables interpretable and strongly generalizable precise anomaly localization and understanding in open-world scenarios.

Abstract

Video Anomaly Detection (VAD) aims to temporally localize abnormal events in videos. Most existing approaches rely on dataset-specific training and curated annotations, limiting generalization in open-set scenarios. Recent zero-shot methods based on Large Vision- Language Models (LVLMs) alleviate this dependency but often lack temporal continuity and structured reasoning. We propose Cog-VADU, a fully training-free framework that reformulates VAD as a sequential cognitive reasoning task. Cog-VADU introduces Chain-of- Anomaly Detection Thought Prompting (CoADTP), which unrolls an LVLM into a recurrent reasoning chain across video segments. By propagating structured rationales over time, the model maintains implicit temporal memory, enabling robust discrimination between com- plex anomalies and high-motion normal activities. To improve reliability, we further design a cross-modal re-ranking stage that aligns textual rationales with visual embeddings, enforcing semantic consistency and temporal coherence for refined and stable predictions. Extensive experiments on multiple public VAD benchmarks demonstrate that Cog-VADU achieves competitive zero-shot performance. Moreover, cross-model evaluations show that CoADTP consistently enhances reasoning-based anomaly detection in a model-agnostic manner, pro- viding interpretable and generalizable anomaly understanding for real-world applications.

Problem

• Video Anomaly Detection
• Zero-shot Generalization
• Open-set Scenarios
• Temporal Continuity
• Structured Reasoning

Innovation

• Training-Free
• Cognitive Reasoning
• Video Anomaly Detection
• Chain-of-Thought Prompting
• Cross-Modal Re-ranking