Investigations on exemplar-based features for speech recognition towards thousands of hours of unsupervised, noisy data.
Georg HeigoldPatrick NguyenMitchel WeintraubVincent VanhouckePublished in: ICASSP (2012)
Keyphrases
- speech recognition
- noisy data
- speech recognition systems
- language model
- hidden markov models
- speech processing
- cepstral coefficients
- speech recognizer
- speech synthesis
- automatic speech recognition
- high dimensional
- speaker identification
- missing data
- high dimensionality
- speech signal
- audio visual speech recognition
- pattern recognition
- low level
- feature extraction
- speech retrieval
- noisy environments
- extracting features
- speaker independent
- speech recognition technology
- computer vision
- information retrieval
- feature set
- feature vectors
- training data
- speaker dependent