Integrating meta-information into exemplar-based speech recognition with segmental conditional random fields.
Kris DemuynckDino SeppiDirk Van CompernollePatrick NguyenGeoffrey ZweigPublished in: ICASSP (2011)
Keyphrases
- speech recognition
- conditional random fields
- hidden markov models
- meta information
- sequence labeling
- speech recognizer
- metadata
- speech synthesis
- random fields
- speech signal
- crf model
- maximum entropy
- named entity recognition
- digital resources
- automatic speech recognition
- graphical models
- speech recognition systems
- speaker identification
- language model
- information extraction
- pattern recognition
- markov random field
- data sources
- natural language