SEMBED: Semantic Embedding of Egocentric Action Videos.
Michael WrayDavide MoltisantiWalterio W. Mayol-CuevasDima DamenPublished in: CoRR (2016)
Keyphrases
- human actions
- action descriptions
- video event
- activity recognition
- sports video
- recognizing human actions
- action classification
- action recognition
- semantic web
- video data
- semantic concepts
- video analysis
- web service descriptions
- recognizing actions
- video summarization
- semantic knowledge
- high level
- video frames
- natural language
- semantic similarity
- semantic representation
- video sequences
- semantic search
- action selection
- video clips
- video database
- video content
- semantic annotation
- spatio temporal
- dynamic scenes
- semantic content
- video surveillance
- human motion
- semantic network
- textual descriptions
- human activities
- action detection
- domain specific
- co occurrence
- image sequences