Detecting abnormality with separated foreground and background: Mutual Generative Adversarial Networks for video abnormal event detection.
Zhi ZhangSheng-hua ZhongAhmed FaresYan LiuPublished in: Comput. Vis. Image Underst. (2022)
Keyphrases
- foreground and background
- abnormal event detection
- foreground objects
- video sequences
- abnormal events
- video frames
- video data
- mean shift
- multimedia
- generative model
- bayesian framework
- background subtraction
- video surveillance
- temporal coherence
- motion segmentation
- color distribution
- spatial information
- moving camera
- pixel wise