Condensing a Sequence to One Informative Frame for Video Recognition.
Zhaofan QiuTing YaoYan ShuChong-Wah NgoTao MeiPublished in: CoRR (2022)
Keyphrases
- video frames
- key frames
- human activities
- successive frames
- video data
- video sequences
- temporal coherence
- single frame
- video content
- pattern recognition
- multimedia
- recognition accuracy
- image frames
- recognition rate
- static images
- input video
- space time
- object recognition
- video retrieval
- automatic recognition
- real time
- neighboring frames
- feature extraction
- activity recognition
- video streams
- temporal continuity
- temporal structure
- video clips
- video signals
- video images
- face recognition
- text detection
- temporal correlation
- multimedia data
- video database
- character recognition
- frame rate
- recognition algorithm
- video shots
- action recognition
- video analysis
- moving objects
- hand gestures
- dynamic scenes