Play estimation with motions and textures with automatic generation of template space-time map.
Kyota AokiRyo AitaTakuro FukibaPublished in: ICDIP (2015)
Keyphrases
- space time
- video sequences
- dynamic scenes
- motion patterns
- spatio temporal
- spatial and temporal
- image sequences
- scan statistic
- human motion
- maximum a posteriori
- video representation
- dynamic textures
- template matching
- super resolution reconstruction
- map estimation
- texture features
- machine learning
- input video
- moving objects
- human actions
- feature vectors
- frame rate
- motion estimation
- maximum likelihood
- multiple view geometry
- video data