Spectral Feature Mapping with MIMIC Loss for Robust Speech Recognition.
Deblin BagchiPeter PlantingaAdam StiffEric Fosler-LussierPublished in: ICASSP (2018)
Keyphrases
- speech recognition
- feature mapping
- noisy environments
- hidden markov models
- speech synthesis
- cepstral coefficients
- speech signal
- pattern recognition
- speech recognizer
- language model
- automatic speech recognition
- speaker adaptation
- speech recognition technology
- speaker identification
- speech recognition systems
- linear prediction
- multi band
- multiple instance learning
- active learning
- image retrieval