Multimodal speaker diarization using oriented optical flow histograms.
Mary Tai KnoxGerald FriedlandPublished in: INTERSPEECH (2010)
Keyphrases
- speaker diarization
- optical flow
- speech recognition
- image sequences
- broadcast news
- multi modal
- audio stream
- bayesian information criterion
- language model
- computer vision
- moving objects
- multimedia
- hidden markov models
- generative model
- artificial neural networks
- gaussian mixture model
- speaker identification
- image processing