Wav2vec-S: Semi-Supervised Pre-Training for Low-Resource ASR.
Han ZhuLi WangGaofeng ChengJindong WangPengyuan ZhangYonghong YanPublished in: INTERSPEECH (2022)
Keyphrases
- semi supervised
- fully labeled
- supervised learning
- training process
- test set
- semi supervised learning
- training set
- active learning
- semi supervised classification
- high levels
- resource allocation
- labeled data
- automatic speech recognition
- speech recognition
- learning algorithm
- batch mode
- training phase
- clustering algorithm
- reinforcement learning
- unlabeled data
- data sets
- multi view
- high dimensional