Directly Locating Actions in Video with Single Frame Annotation.
Haoran TongXinyan LiuGuorong LiLaiyun QingPublished in: ICMR (2024)
Keyphrases
- single frame
- video frames
- video sequences
- super resolution
- high resolution video
- multiple frames
- image sequences
- multi frame
- optical flow
- surveillance videos
- low resolution
- human actions
- metadata
- video data
- key frames
- image annotation
- high resolution
- viewpoint
- temporal coherence
- human computer interaction
- video content
- object detection
- video analysis
- facial expressions
- image retrieval
- temporal redundancy
- multimedia