SEMBED: Semantic Embedding of Egocentric Action Videos.
Michael WrayDavide MoltisantiWalterio W. Mayol-CuevasDima DamenPublished in: ECCV Workshops (1) (2016)
Keyphrases
- human actions
- action descriptions
- action recognition
- video event
- activity recognition
- video sequences
- sports video
- high level
- recognizing human actions
- action classification
- video frames
- video summarization
- web service descriptions
- recognition of human actions
- semantic search
- video data
- human activities
- view invariant
- natural language
- spatio temporal
- semantic similarity
- computer vision
- recognizing actions
- action detection
- video database
- video clips
- semantic network
- vector space
- textual descriptions
- space time
- video content
- data hiding
- surveillance videos
- low dimensional
- video analysis