Integer linear programming for speaker diarization and cross-modal identification in TV broadcast.
Hervé BredinJohann PoignantPublished in: INTERSPEECH (2013)
Keyphrases
- integer linear programming
- cross modal
- speaker diarization
- multi modal
- tv broadcast
- column generation
- multimedia retrieval
- multimedia databases
- image retrieval
- global constraints
- visual similarity
- speech recognition
- audio visual
- cutting plane
- image search
- visual data
- e learning
- linear programming
- broadcast news
- pattern recognition