Text Prompt with Normality Guidance for Weakly Supervised Video Anomaly Detection

arXiv:2404.08531v1 Announce Type: new
Abstract: Weakly supervised video anomaly detection (WSVAD) is a challenging task. Generating fine-grained pseudo-labels based on weak-label and then self-training a classifier is currently a promising solution. However, since the existing methods use only RGB visual modality and the utilization of category text information is neglected, thus limiting the generation of more accurate pseudo-labels and affecting the performance of self-training. Inspired by the manual labeling process based on the event description, in this paper, we propose …

abstract anomaly anomaly detection arxiv classifier cs.cv detection fine-grained guidance however information labels normality prompt self-training solution text training type video visual