Jurnal Inovasi Teknologi dan Rekayasa ISSN: 2581-1274 Vol. 4, No 1, June 2019, pp. 31-31 31 DOI: 10.31572/inotera.Vol4.Iss1.2019.ID73 W: http://inotera.poltas.ac.id | E: inotera@poltas.ac.id Feature Extraction of Facial Electromyography (EMG) Signal for Aceh Languages Speech using Discrete Wavelet Transform (DWT) Darma Setiawan Putra a,1,* , Yuril Umbu Woza Weru b,2 a,b Informatics Engineering, Politeknik Aceh Selatan, Merdeka Street, Tapaktuan, 23711, Indonesia 1 darma.poltas@gmail.com*; 2 yurilweru@gmail.com * corresponding author I. Introduction The development of science and technology has significantly influenced in signal processing technology. The human has electromyography (EMG) signal in the arm, leg, and facial muscle. The EMG signal in the face is called the facial electromyography (FEMG) signal. The FEMG is usually used to measure the level of emotional expression in a human face. This FEMG signal is recorded by attaching a surface electrode to the skin of the face to find out the emotional expression. The emotional expressions that are usually measured are expression of pleasure, sadness, and anger. The facial muscles used to measure the EMG signals are corrugator supercili muscle, zygomaticus major and orbicularis oculi [1][2][3]. This emotional expression is interpreted by measuring the magnitude of facial EMG signals. Emotional expression can also be associated with specific FEMG signal pattern. The FEMG signal can also show physiological conditions as an objective measure of facial muscle activity. When there is a change in facial expression, it will cause contractions in the facial muscle. This FEMG signal can be used to recognize human facial gesture for assistive technology and rehabilitation [4][5][6]. In addition, measuring emotional expression and facial gesture, this FEMG signal can measure signal pattern when the human speech in daily communication [7][8][9][10]. When human speaks the languages, the articulation muscles around the mouth contracts that can cause FEMG signal. This study aims to analyze the feature of the FEMG signal when speaking in Aceh language. So that this feature can be used for pattern recognition of FEMG signal in human facial muscle. ARTICLE INFO ABSTRACT Article history: Accepted The facial electromyography (FEMG) signal is a signal that occurs in the muscles of the contracted human face. This FEMG signal is one of the techniques used to study human speech recognition. It can be acquired by placing an electrode surface on the skin around the facial articulation muscle. Three types of muscles in this study are the masseter, rigorous and depressor muscle. This study aims to extract and analyze the features in the FEMG signal. The extraction method is the discrete wavelet transform (DWT). The type of wavelet transform is Daubechies2 with level 5. After extraction and analysis of FEMG signals, the FEMG signal pattern for each spoken word indicated by differences in the approximation and detail coefficient of the FEMG signal. In addition, the level of difference in the FEMG signal pattern is also indicated by the histogram of the approximation coefficient of the FEMG signal. Thus, the discrete wavelet transform method can be used as one of the methods for extracting the FEMG signal feature in a human facial electromyography (FEMG) signal. Copyright © 2019 Politeknik Aceh Selatan. All rights reserved. Keywords: Facial electromyography FEMG Speech recognition Discrete wavelet transform Feature extraction