The effects of different training modes on the performance of silent speech recognition based on high-density sEMG.
Yao PiMingxing ZhuZijian YangXin WangCheng WangHaoshi ZhangMingjiang WangFeng WanShixiong ChenGuanglin LiPublished in: RCAR (2021)
Keyphrases
- speech recognition
- high density
- wall street journal corpus
- low density
- isolated word
- acoustic models
- hidden markov models
- language model
- speech recognizer
- speech synthesis
- speech processing
- speech understanding
- noisy environments
- pattern recognition
- speech signal
- speech recognition systems
- automatic speech recognition
- speech recognition technology
- thin film
- data center
- field effect transistors
- speaker identification
- training process
- speaker independent
- magnetic tape
- cepstral coefficients
- keyword spotting
- speech retrieval
- neural network
- background noise
- statistically significant
- energy consumption
- learning styles
- chemical vapor deposition
- speaker dependent
- signal processing
- face recognition
- information retrieval