FAT-HuBERT: Front-End Adaptive Training of Hidden-Unit BERT For Distortion-Invariant Robust Speech Recognition.
Dongning YangWei WangYanmin QianPublished in: ASRU (2023)
Keyphrases
- speech recognition
- noisy environments
- wall street journal corpus
- hidden markov models
- language model
- automatic speech recognition
- speech signal
- speech synthesis
- isolated word
- speech recognizer
- acoustic models
- speaker identification
- speech understanding
- pattern recognition
- speech processing
- training process
- speech recognition systems
- keyword spotting
- cepstral coefficients
- speech recognition errors
- speech recognition technology
- speech recognizers
- speaker diarization
- neural network
- speaker recognition
- training set