MKD: Mixup-Based Knowledge Distillation for Mandarin End-to-End Speech Recognition.
Xing WuYifan JinJianjia WangQuan QianYike GuoPublished in: Algorithms (2022)
Keyphrases
- speech recognition
- end to end
- speech signal
- hidden markov models
- speech synthesis
- automatic speech recognition
- speech recognition technology
- language model
- speech processing
- speaker independent
- pattern recognition
- speech recognizer
- congestion control
- speaker identification
- speaker dependent
- acoustic models
- isolated word
- speech recognition systems
- noisy environments
- natural language processing
- feature vectors