LatticeBART: Lattice-to-Lattice Pre-Training for Speech Recognition.
Lingfeng DaiLu ChenZhikai ZhouKai YuPublished in: ICASSP (2022)
Keyphrases
- speech recognition
- wall street journal corpus
- hidden markov models
- automatic speech recognition
- speech signal
- speech synthesis
- isolated word
- speaker identification
- speech processing
- language model
- pattern recognition
- speech understanding
- training set
- keyword spotting
- speech recognizer
- speaker recognition
- speech recognition technology
- acoustic models
- speech recognizers
- noisy environments
- audio visual speech recognition
- information retrieval
- speaker diarization
- machine learning
- speaker dependent
- speech retrieval