Deep Image-to-Video Adaptation and Fusion Networks for Action Recognition.
Yang LiuZhaoyang LuJing LiTao YangChao YaoPublished in: CoRR (2019)
Keyphrases
- action recognition
- static images
- human actions
- action classification
- mid level
- bag of features
- image classification
- spatial temporal
- multiscale
- computer vision
- video dataset
- action detection
- bag of words
- image features
- image content
- activity recognition
- video sequences
- recognizing human actions
- image representation
- recognition of human actions
- high resolution
- motion features
- global features
- space time interest points
- human detection
- body parts
- video frames
- video images
- human activities
- visual features
- view invariant
- input image
- low level
- image retrieval
- object recognition
- class specific
- recognizing actions
- low level descriptors