VQ-VAD: Vector-quantized Motion Representation Learning for Human-centric Video Anomaly Detection

作者：Narges Rashvand, Ghazal Alinezhad Noghre, Shanle Yao, Gabriel Maldonado, Hamed Tabkhi · 单位：Electrical and Computer Engineering Department, University of North Carolina at Charlotte, base for this work is available at https://github.com/TeCSAR-UNCC/VQ-VAD., anomalies. While these approaches have advanced the field by reducing the need for labeled anomalies, they · 会议/期刊：arXiv preprint · 方向：cs.CV · 发布日期：2026-08-06