Comparing Self-Supervised Pre-Training and Semi-Supervised Training for Speech Recognition in Languages with Weak Language Models.
Léa-Marie Lam-Yee-MuiLucas Ondel YangOndrej KlejchPublished in: INTERSPEECH (2023)
Keyphrases
- speech recognition
- language model
- supervised training
- training phase
- language modeling
- conditional random fields
- training algorithm
- probabilistic model
- unsupervised learning
- document retrieval
- supervised learning
- query expansion
- cross lingual
- n gram
- retrieval model
- speech signal
- query terms
- hidden markov models
- automatic speech recognition
- information retrieval
- handwriting recognition
- test collection
- language independent
- dependency parsing
- neural network
- relevance model
- training set
- query translation
- translation model
- graphical models
- feature extraction
- out of vocabulary