An evaluation of unsupervised acoustic model training for a dysarthric speech interface.
Oliver WalterVladimir DespotovicReinhold Haeb-UmbachJort F. GemmekeBart OnsHugo Van hammePublished in: INTERSPEECH (2014)
Keyphrases
- supervised learning
- speech recognition
- hearing impaired
- supervised training
- pattern recognition
- training set
- hands free
- user interface
- natural language processing
- machine learning
- training samples
- evaluation method
- training algorithm
- unsupervised learning
- direct manipulation
- training phase
- speech signal
- data driven
- hidden markov models
- pairwise
- search engine
- information retrieval