Motion = Video - Content: Towards Unsupervised Learning of Motion Representation from Videos.
Hehe FanMohan S. KankanhalliPublished in: MMAsia (2021)
Keyphrases
- video content
- video representation
- key frames
- unsupervised learning
- space time
- video sequences
- video footage
- video data
- motion estimation
- video streams
- video objects
- video summarization
- image sequences
- optical flow
- video quality
- spatio temporal
- video analysis
- video retrieval
- video clips
- video frames
- video database
- motion patterns
- video shots
- dynamic scenes
- human motion
- three dimensional
- video classification
- dynamic textures
- machine learning
- video signals
- video browsing
- video collections
- interesting events
- content based video
- low level features
- video segments
- supervised learning
- sports video
- video annotation
- moving objects
- object recognition