Experiments on speech tracking in audio documents using Gaussian mixture modeling.
Mouhamadou SeckIvan Magrin-ChagnolleauFrédéric BimbotPublished in: ICASSP (2001)
Keyphrases
- gaussian mixture modeling
- spoken documents
- audio visual
- audio stream
- broadcast news
- speaker identification
- automatic transcription
- gaussian mixture model
- information retrieval
- audio features
- visual information
- image representation
- particle filter
- multimedia
- real time
- automatic speech recognition
- mean shift
- video streams
- image data
- probabilistic model
- visual speech
- image sequences
- feature selection