“…Therefore, it can be used even if the subject does not produce any audible voices (silent speech), such as patients with speech disorders [2,3]. Therefore, the sEMG-based ASR system has developed into a prevalent technique with a wide variety of applications for speaking recognition in both audible and silent modes [4][5][6].…”