An evaluation of cross-language adaptation and native speech training for rapid HMM construction based on very limited training data.
Xufang ZhaoDouglas D. O'ShaughnessyPublished in: INTERSPEECH (2007)
Keyphrases
- cross language
- speaker adaptation
- speech recognition
- speech retrieval
- acoustic models
- automatic speech recognition
- training data
- hidden markov models
- spoken document retrieval
- training set
- speech signal
- text retrieval
- speech recognizer
- language independent
- speaker independent
- supervised learning
- document retrieval
- question answering
- cross language information retrieval
- cross lingual
- data sets
- text categorization
- document collections
- prior knowledge
- broadcast news
- learning algorithm
- neural network
- information access
- speaker identification
- spontaneous speech
- conversational speech
- language model
- text classification
- nearest neighbor
- cl sr