ViTframe: Vision Transformer Acceleration via Informative Frame Selection for Video Recognition.
Chunyu QiZilong LiZhuoran SongXiaoyao LiangPublished in: ICCD (2023)
Keyphrases
- video frames
- recognition rate
- real time
- object recognition
- successive frames
- temporal coherence
- image frames
- video sequences
- single frame
- video data
- recognition accuracy
- computer vision
- video content
- multimedia
- video clips
- key frames
- human activities
- pattern recognition
- visual perception
- feature extraction
- spatial and temporal
- video streams
- activity recognition
- power system
- input video
- video surveillance
- vision system
- fuzzy logic
- video analysis
- video signals
- visual processing
- video images
- recognition process
- temporal continuity
- dynamic scenes
- recognition algorithm
- multimedia data
- space time
- human computer interaction
- neural network
- video database
- character recognition
- video retrieval