ST-Adapter: Parameter-Efficient Image-to-Video Transfer Learning for Action Recognition.
Junting PanZiyi LinXiatian ZhuJing ShaoHongsheng LiPublished in: CoRR (2022)
Keyphrases
- action recognition
- static images
- transfer learning
- human actions
- action classification
- activity recognition
- input image
- image features
- learning tasks
- recognition of human actions
- recognizing human actions
- human activities
- bag of words
- action detection
- computer vision
- video frames
- reinforcement learning
- image segmentation
- body parts
- image representation
- labeled data
- image classification
- video sequences
- high resolution
- active learning
- visual concepts
- machine learning
- video content
- machine learning algorithms
- text mining
- image collections
- multimedia
- similarity measure
- motion history images
- semi supervised learning