MoLo: Motion-Augmented Long-Short Contrastive Learning for Few-Shot Action Recognition.
Xiang WangShiwei ZhangZhiwu QingChangxin GaoYingya ZhangDeli ZhaoNong SangPublished in: CVPR (2023)
Keyphrases
- action recognition
- static images
- recognition of human actions
- computer vision
- activity recognition
- human actions
- recognizing human actions
- bag of words
- action recognition in videos
- key frames
- motion field
- human motion
- max margin
- motion features
- view invariant
- machine learning
- body parts
- bag of features
- human detection
- motion estimation
- image sequences