Unsupervised Anomaly Video Detection via a Double-Flow ConvLSTM Variational Autoencoder.
Lin WangHaishu TanFuqiang ZhouWangxia ZuoPengfei SunPublished in: IEEE Access (2022)
Keyphrases
- anomaly detection
- face detection and tracking
- video sequences
- activity detection
- event detection
- multimedia
- automatic detection
- detection method
- false alarms
- detection algorithm
- object detection and tracking
- unsupervised manner
- video database
- video data
- space time
- multi modal fusion
- real time
- video scene
- false positives
- video streams
- text detection
- soccer video
- video segmentation
- semi supervised
- detection rate
- video frames
- video clips
- video retrieval
- video content
- unsupervised learning
- intrusion detection
- detect anomalies
- person detection
- spatial and temporal
- flow field
- surveillance videos
- restricted boltzmann machine
- shot boundary detection
- motion estimation