FFNeRV: Flow-Guided Frame-Wise Neural Representations for Videos.
Joo Chan LeeDaniel RhoJong Hwan KoEunbyung ParkPublished in: ACM Multimedia (2023)
Keyphrases
- video frames
- neighboring frames
- successive frames
- temporal coherence
- distributed representations
- key frames
- video sequences
- input video
- neural network
- network architecture
- adjacent frames
- video data
- consecutive frames
- video content
- internal representations
- flow field
- associative memory
- frame rate
- reference frame
- bio inspired
- video objects
- neural model
- multiple representations
- pairwise
- rigidly moving
- video summarization
- visual patterns
- video analysis
- dynamic scenes
- higher level
- spatio temporal
- computer vision