Predicting positive transfer for improved low-resource speech recognition using acoustic pseudo-tokens.
Nay SanGeorgios ParaskevopoulosAryaman AroraXiluo HePrabhjot KaurOliver AdamsDan JurafskyPublished in: CoRR (2024)
Keyphrases
- automatic speech recognition
- speech sounds
- acoustic features
- speech recognition
- speech signal
- speech recognizers
- speech recognition systems
- noisy environments
- positive and negative
- speaker independent
- resource allocation
- information retrieval
- transfer learning
- resource management
- improved algorithm
- spectral features
- positively correlated
- vocal tract
- acoustic signal
- feature extraction