FROSTER: Frozen CLIP is A Strong Teacher for Open-Vocabulary Action Recognition.
Xiaohu HuangHao ZhouKun YaoKai HanPublished in: ICLR (2024)
Keyphrases
- action recognition
- human actions
- bag of words
- human detection
- computer vision
- action classification
- activity recognition
- spatial temporal
- body parts
- recognizing human actions
- human pose
- human activities
- video dataset
- recognition of human actions
- recognizing actions
- mid level
- static images
- view invariant action recognition
- independent subspace analysis
- view invariant
- max margin
- object recognition