Multi-stream spectro-temporal and cepstral features based on data-driven hierarchical phoneme clusters.
Shang-wen LiLiang-Che SunLin-Shan LeePublished in: ICASSP (2011)
Keyphrases
- data driven
- multi stream
- cepstral features
- hidden markov models
- hierarchical clustering
- audio visual speech recognition
- hierarchical structure
- audio visual
- clustering algorithm
- speech recognition
- spatio temporal
- feature extraction
- contextual information
- multi modal
- automatic speech recognition
- data analysis
- keywords
- image content