Improving Transformer-Based Speech Recognition with Unsupervised Pre-Training and Multi-Task Semantic Knowledge Learning.
Song LiLin LiQingyang HongLingling LiuPublished in: INTERSPEECH (2020)
Keyphrases
- speech recognition
- multi task
- learning tasks
- semantic knowledge
- supervised learning
- multi task learning
- learning problems
- hidden markov models
- learning process
- unsupervised learning
- reinforcement learning
- knowledge acquisition
- language model
- training set
- learning algorithm
- transfer learning
- prior knowledge
- pattern recognition
- training data