Large synthetic dataset improves fine-grained video anomaly understanding
AnomalyCraft-700K: Component-Level Controllable and Verifiable Synthetic Anomalies for Fine-Grained Video Anomaly Understanding
Computer Vision and Pattern Recognition
Summary
Video anomaly detection is hard because real unusual events are rare and difficult to collect with good details. The authors created a huge synthetic video dataset called AnomalyCraft-700K, with detailed, controllable anomalies and precise text descriptions that match video components. This dataset helps train models to better recognize subtle anomalies and understand them in fine detail. They also included tricky normal examples that look similar to anomalies to make the detection smarter.
What this means in practice
- •For security monitoring teams: Use the dataset to train surveillance systems that detect subtle and rare abnormal events more accurately, improving safety and response.
- •For video analytics developers: Build fine-grained anomaly recognition tools that understand complex event details and generate descriptive captions or retrieve relevant anomaly clips.
Authors
Yuzhou Long, Haodong Zhang, Yunpeng Yang, Peng Wu, Guansong Pang
Abstract
Progress in video anomaly understanding (VAU) has long been limited by inherent deficiencies of real-world anomaly videos, which are hard to collect and offer little control over their content. Synthetic anomaly approaches partially alleviate data scarcity, yet their generation remains largely controlled at the category or prompt level. They also lack component-level verification of video-text consistency and provide insufficient hard normal samples near the normal-anomaly boundary. To address this, we present AnomalyCraft-700K, a component-controllable synthetic anomaly dataset for fine-grained VAU, containing over 40K videos and over 700K task-level textual annotations. From fine-grained semantic components and a progressive three-stage pipeline, we craft anomaly events that are richly detailed, semantically controlled, and temporally structured, and additionally construct per-category hard normal samples to prompt the model to discriminate based on anomaly semantics rather than surface visual cues. Moreover, using the components as verification units, AnomalyCraft-700K further performs component-wise correction of video-text discrepancies introduced during generation, providing reliable annotations with verified cross-modal alignment for six tasks that progress from anomaly detection, through anomaly retrieval and captioning, to fine-grained anomaly reasoning. Evaluations of widely used methods under both traditional and MLLM-based protocols demonstrate that AnomalyCraft-700K serves as an effective source of supervision, from anomaly detection to fine-grained anomaly understanding.