EZ-CLIP: Efficient Zeroshot Video Action Recognition.
Shahzad AhmadSukalpa ChandaYogesh S. RawatPublished in: CoRR (2023)
Keyphrases
- action recognition
- human actions
- action classification
- spatial temporal
- video dataset
- action detection
- recognizing human actions
- static images
- recognition of human actions
- human activities
- computer vision
- activity recognition
- video clips
- spatio temporal interest points
- human detection
- bag of words
- mid level
- space time interest points
- motion features
- event recognition
- depth sensors
- space time
- video retrieval
- video data
- view invariant
- bag of features
- recognizing actions
- action primitives
- view invariant action recognition
- depth information
- body parts
- event detection
- object recognition
- high level