A Weighted Sparse Sampling and Smoothing Frame Transition Approach for Semantic Fast-Forward First-Person Videos.
Michel Melo SilvaWashington Luis Souza RamosJoão Pedro Klock FerreiraFelipe Cadar ChamoneMario Fernando Montenegro CamposErickson Rangel do NascimentoPublished in: CoRR (2018)
Keyphrases
- sparse sampling
- video frames
- video event
- temporal coherence
- key frames
- neighboring frames
- natural language
- semantic concepts
- video sequences
- domain specific
- semantic similarity
- input video
- semantic search
- video content
- low level features
- semantic video retrieval
- adjacent frames
- semantic web
- sports video
- semantic representation
- human activities
- image frames
- semantic information
- video objects
- video analysis
- high level
- single frame
- video clips
- successive frames
- event detection
- dynamic scenes
- smoothing methods
- semantic network
- video data
- text classification
- multiscale