CMAE-V: Contrastive Masked Autoencoders for Video Action Recognition.
Chengze LuXiaojie JinZhicheng HuangQibin HouMing-Ming ChengJiashi FengPublished in: CoRR (2023)
Keyphrases
- action recognition
- human actions
- action classification
- spatial temporal
- video dataset
- action detection
- recognizing human actions
- recognition of human actions
- motion features
- human activities
- static images
- bag of words
- activity recognition
- spatio temporal interest points
- space time interest points
- denoising
- computer vision
- video sequences
- body parts
- human detection
- mid level
- recognizing actions
- multimedia
- video data
- motion history images
- video analysis
- video streams
- video images
- bag of features
- view invariant
- video content
- video frames
- action recognition in videos
- view invariant action recognition
- event recognition
- human motion
- action primitives
- space time