Unsupervised Learning of View-invariant Action Representations.
Junnan LiYongkang WongQi ZhaoMohan S. KankanhalliPublished in: CoRR (2018)
Keyphrases
- view invariant
- unsupervised learning
- action recognition
- human actions
- multi view
- human motion analysis
- single view
- human motion
- semi supervised
- supervised learning
- viewpoint
- cross view action recognition
- activity recognition
- dimensionality reduction
- spatio temporal
- space time
- human activities
- visual features
- gesture recognition
- invariant representation
- invariant representations
- human body
- multiple views
- training data
- expectation maximization
- text classification
- data points
- feature space
- object recognition