Learning motion representation for real-time spatio-temporal action localization.
Dejun ZhangLinchao HeZhigang TuShifu ZhangFei HanBoxiong YangPublished in: Pattern Recognit. (2020)
Keyphrases
- real time
- spatio temporal
- image sequences
- learning process
- spatial and temporal
- recognizing human actions
- supervised learning
- human actions
- recognition of human actions
- temporal domain
- spatio temporal data
- motion analysis
- learning tasks
- space time
- learning systems
- prior knowledge
- learning algorithm
- action recognition
- point correspondences
- online learning
- relative position
- three dimensional
- multiple representations
- action models
- view invariant
- computer vision