Decoding silent speech from high-density surface electromyographic data using transformer

doi:10.1016/j.bspc.2022.104298

	Decoding silent speech from high-density surface electromyographic data using transformer
	Song, Rui 1; Zhang, Xu 1; Chen, Xi 1; Chen, Xiang 1; Chen, Xun 1; Yang, Shuang 2; Yin, Erwei 3
	2023-02-01
发表期刊	BIOMEDICAL SIGNAL PROCESSING AND CONTROL
ISSN	1746-8094
卷号	80 页码:9
摘要	Recent silent speech recognition (SSR) studies based on surface electromyography (sEMG) have been conducted by classifying a finite number of words or phrases without sufficient understanding of temporally semantic in-formation compared to sequential decoding at a fine-grained syllable or phoneme level. This paper presents a syllable-level sequential decoding method using a transformer model for sEMG-based SSR. The proposed method consists of a transformer model and a language model. The input sEMG data was first translated into a sequence of syllable-level decisions by the transformer model. Then, these sequential syllable-level decisions were tuned as a final syllable sequence to approximate natural language through the language model. To verify the effec-tiveness of the proposed method, experiment data were recorded using two high-density electrode arrays with 64 channels from a total of eight subjects during subvocally reading a corpus of 33 Chinese phrases generated from a dictionary of 82 syllables. The proposed method achieved the lowest character error rate of 5.14 +/- 3.28 % and the highest phrase recognition accuracy of 96.37 +/- 2.06 %, and it significantly outperformed other common methods for sEMG-based SSR. These findings demonstrated the feasibility and usability of the proposed method for practical SSR applications.
关键词	Surface electromyography Silent speech recognition Sequential decoding Transformer Language model
DOI	10.1016/j.bspc.2022.104298
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China[62271464] ; National Natural Science Foundation of China[62076250]
WOS研究方向	Engineering
WOS类目	Engineering, Biomedical
WOS记录号	WOS:000890503700003
出版者	ELSEVIER SCI LTD
引用统计	被引频次：5[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/20270
专题	中国科学院计算技术研究所期刊论文
通讯作者	Zhang, Xu
作者单位	1.Technol Univ Sci & Technol China, Sch Informat Sci, Hefei, Anhui, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, Beijing, Peoples R China 3.Acad Mil Sci Peoples Liberat Army, Natl Innovat Inst Def Technol, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Song, Rui,Zhang, Xu,Chen, Xi,et al. Decoding silent speech from high-density surface electromyographic data using transformer[J]. BIOMEDICAL SIGNAL PROCESSING AND CONTROL,2023,80:9.
APA	Song, Rui.,Zhang, Xu.,Chen, Xi.,Chen, Xiang.,Chen, Xun.,...&Yin, Erwei.(2023).Decoding silent speech from high-density surface electromyographic data using transformer.BIOMEDICAL SIGNAL PROCESSING AND CONTROL,80,9.
MLA	Song, Rui,et al."Decoding silent speech from high-density surface electromyographic data using transformer".BIOMEDICAL SIGNAL PROCESSING AND CONTROL 80(2023):9.