Entanglement Loss for Context-Based Still Image Action Recognition.
Miao XinShuhang WangJian ChengPublished in: ICME (2019)
Keyphrases
- action recognition
- static images
- mid level
- bag of features
- bag of words
- human actions
- input image
- image classification
- image representation
- action classification
- image features
- single image
- activity recognition
- image content
- body parts
- human detection
- computer vision
- recognition of human actions
- multiscale
- image segmentation
- keypoints
- image retrieval
- view invariant
- human activities
- segmentation method
- high resolution
- spatial information
- context sensitive
- partial occlusion
- human body
- action detection
- depth sensors
- recognizing human actions
- recognizing actions