FFNeRV: Flow-Guided Frame-Wise Neural Representations for Videos.
Joochan LeeDaniel RhoJong Hwan KoEunbyung ParkPublished in: CoRR (2022)
Keyphrases
- video frames
- neighboring frames
- video sequences
- key frames
- temporal coherence
- distributed representations
- successive frames
- neural network
- video data
- network architecture
- adjacent frames
- video content
- video database
- video stabilization
- neural model
- consecutive frames
- flow field
- video analysis
- pairwise
- flow patterns
- visual features
- human activities
- video clips
- single frame
- event recognition
- reference frame
- video dataset
- multiple representations
- input video
- video objects
- stereoscopic video
- artificial neural networks
- multimedia