Speaker Identification by Multi-Frame Generative Models.
Donato ImpedovoMario ReficePublished in: IAS (2008)
Keyphrases
- speaker identification
- multi frame
- generative model
- gaussian mixture model
- mixture model
- motion estimation
- speech recognition
- image sequences
- point correspondences
- em algorithm
- probabilistic model
- noisy environments
- speech signal
- super resolution
- optical flow
- filtering algorithm
- feature extraction
- motion analysis
- prior knowledge
- expectation maximization
- optic flow
- topic models
- broadcast news
- image classification
- feature selection
- video frames
- text mining
- semi supervised
- low level