Exploiting Completeness and Uncertainty of Pseudo Labels for Weakly Supervised Video Anomaly Detection.
Chen ZhangGuorong LiYuankai QiShuhui WangLaiyun QingQingming HuangMing-Hsuan YangPublished in: CoRR (2022)
Keyphrases
- anomaly detection
- weakly labeled
- weakly supervised
- fully supervised
- intrusion detection
- video sequences
- topic models
- relation extraction
- object class
- superpixels
- unsupervised learning
- semi supervised
- training data
- data sets
- named entities
- object detectors
- natural images
- automatic extraction
- multiscale
- detect anomalies
- computer vision