Summary
Real videos of unusual or abnormal events are hard to collect and control, which makes teaching computers to understand them difficult. To fix this, the authors created AnomalyCraft-700K, a large set of computer-made videos showing detailed and controlled abnormal events with clear text descriptions. They also made sure the videos and descriptions match closely by checking each part of the videos and fixing any mistakes. This dataset helps computers learn not only to spot anomalies but also to explain and reason about them in detail. Tests show the dataset works well for improving several video anomaly understanding tasks.
video anomaly understandingsynthetic datasetsfine-grained annotationsvideo-text alignmentanomaly detectionanomaly retrievalanomaly captioningcomponent-level controlhard normal samples
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.