Iterative pseudo-forced alignment by acoustic CTC loss for self-supervised ASR domain adaptation.
Fernando LópezJordi LuquePublished in: IberSPEECH (2022)
Keyphrases
- domain adaptation
- manifold alignment
- multiple sources
- cross domain
- transfer learning
- semi supervised learning
- target domain
- test data
- semi supervised
- labeled data
- automatic speech recognition
- machine learning
- unlabeled data
- data sets
- knn
- training data
- document classification
- sentiment classification
- co training
- test set
- speech recognition
- text categorization
- k nearest neighbor
- data mining
- databases