Exploring Self-Supervised Pre-Trained ASR Models for Dysarthric and Elderly Speech Recognition.
Shujie HuXurong XieZengrui JinMengzhe GengYi WangMingyu CuiJiajun DengXunying LiuHelen MengPublished in: ICASSP (2023)
Keyphrases
- speech recognition
- automatic speech recognition
- acoustic models
- language model
- hidden markov models
- speech recognizer
- noisy environments
- pre trained
- speech synthesis
- speech signal
- word error rate
- speech retrieval
- probabilistic model
- speaker identification
- pattern recognition
- neural network
- speech recognition systems
- speech recognition technology
- speaker independent
- support vector