REST: REtrieve & Self-Train for generative action recognition.
Adrian BulatEnrique SanchezBrais MartínezGeorgios TzimiropoulosPublished in: CoRR (2022)
Keyphrases
- action recognition
- human actions
- bag of words
- activity recognition
- human detection
- computer vision
- action classification
- generative model
- spatial temporal
- recognizing human actions
- static images
- bag of features
- video dataset
- view invariant action recognition
- recognizing actions
- mid level
- human activities
- human pose
- action recognition in videos
- multi class
- body parts
- view invariant
- particle filter
- recognition of human actions
- independent subspace analysis