Using Deep-Q Network to Select Candidates from N-best Speech Recognition Hypotheses for Enhancing Dialogue State Tracking.
Richard Tzong-Han TsaiChia-Hao ChenChun-Kai WuYu-Cheng HsiaoHung-yi LeePublished in: ICASSP (2019)
Keyphrases
- speech recognition
- hidden markov models
- language model
- speech processing
- audio visual speech recognition
- speech synthesis
- automatic speech recognition
- speech recognizer
- speech understanding
- speaker identification
- speech signal
- speech recognition systems
- noisy environments
- pattern recognition
- speaker independent
- speaker diarization
- neural network
- speech recognition errors
- machine learning
- dialogue system
- natural language processing
- cepstral coefficients
- speech recognition technology
- speech recognizers