DDLSTM: Dual-Domain LSTM for Cross-Dataset Action Recognition.
Toby PerrettDima DamenPublished in: CoRR (2019)
Keyphrases
- action recognition
- human actions
- video dataset
- bag of words
- ucf sports
- action classification
- activity recognition
- human detection
- spatial temporal
- computer vision
- body parts
- space time interest points
- human activities
- action recognition in videos
- recognition of human actions
- recognizing actions
- static images
- human pose
- independent subspace analysis
- recognizing human actions
- mid level
- view invariant
- bag of features
- depth sensors
- spatio temporal
- view invariant action recognition
- action detection
- motion recognition
- human activity recognition
- human computer interaction
- video sequences