Transductive Zero-Shot Action Recognition by Word-Vector Embedding.
Xun XuTimothy M. HospedalesShaogang GongPublished in: Int. J. Comput. Vis. (2017)
Keyphrases
- action recognition
- vector space
- bag of words
- human actions
- activity recognition
- computer vision
- spatial temporal
- body parts
- action classification
- n gram
- human detection
- recognizing human actions
- co occurrence
- semi supervised
- bag of features
- recognition of human actions
- independent subspace analysis
- action primitives
- text classification
- human activities
- static images
- view invariant
- support vector
- recognizing actions
- human pose
- action recognition in videos
- feature vectors
- max margin
- video dataset
- depth cameras
- text categorization
- semi supervised learning
- labeled data
- visual features
- multi view
- text mining
- view invariant action recognition