Rapid Feature Space Speaker Adaptation for Multi-Stream HMM-Based Audio-Visual Speech Recognition.
Jing HuangEtienne MarcheretKarthik VisweswariahPublished in: ICME (2005)
Keyphrases
- speaker adaptation
- audio visual speech recognition
- multi stream
- hidden markov models
- speech recognition
- feature space
- automatic speech recognition
- speaker independent
- speech recognizer
- feature vectors
- feature selection
- high dimensional
- noisy environments
- principal component analysis
- maximum likelihood
- dimensionality reduction
- image retrieval
- data points
- speech signal
- feature set
- speaker identification
- computer vision
- information retrieval
- audio visual
- visual words
- gaussian mixture model
- image representation
- mixture model
- multi modal
- visual speech
- language model
- feature extraction