Computer-based Malay Articulation Training for Malay Plosives at Isolated, Syllable and Word Level H N Ting 1 ; J Yunus 1 ; S Vandort 2 and L C Wong 1 1 Faculty of Electrical Engineering, Universiti Teknologi Malaysia, Malaysia 2 Department of Audiology and Speech Sciences, Universiti Kebangsaan Malaysia, Malaysia Abstract This paper describes the use of computer as an articulation training system for Malay plosives at isolated, syllable and word level. The proposed system is more convenient than the traditional speech analyzing tools such as Electropalatograph, where the latter requires an external electronic circuit to be attached into the mouth of client. The system is designed in a way that is user friendly and easy to use for the speech-language pathologist or even the client. The client undergoes speech training by just talking into the microphone and the system is able to recognize the sounds and classify them accordingly. Audio and visual feedback is used to help the client to identify his or her articulation errors as well as to make comparisons between his/her articulation models with the standard model. The system can be used for both children and adults. 1. Introduction Articulation is basically speech sound production, which is the result from movements of the lips, tongue, velum and jaws to shape the flow of air into sounds [1]. The speech sounds are comprised of three groups: vowels, diphthongs and consonants. Vowels are produced with a relatively open oral cavity with phonation. Diphthongs are vowel-like sounds that are produced as a combination of two vowels. Consonant sounds are produced by constricting the airflow in the oral cavity or the nasal cavity. Articulation problems are concerned with the failure in mastering the speech sound production in a spoken language. The causes of the articulation problems can be organic or functional. The organic causes include anatomical, motor or sensory impairments whereas the functional causes have various etiologies [2]. Consonant sounds are relatively more difficult to be produced than the vowel and diphthong sounds and hence most articulation problems involve consonant sounds. Thus, the research focuses on the articulation of consonant sounds instead of vowel sounds. In conventional speech training, all kinds of material can be used as tools to help treat the articulation problem. These materials include a candle (to detect bilabial sounds), peanut butter (to detect alveolar and velar sounds), a mirror (to provide visual feedback), a tongue depressor (to help cue the client on correct positioning of the tongue), cosmetic mirrors (detect breath vapour), cotton ball (to detect lingua-dental sounds) and so on. The use of these devices is welcomed because they are cheap in cost and the client can produce the speech sounds in a natural manner without any devices connected to his/her mouth. However, this kind of training may require total involvement of a speech-language pathologist (SLP) to monitor and evaluate the client’s speech production performance, where the job is quite laborious and repetitious. Special speech devices such as the Electropalatograph (EPG) and spectrograph are now used in analyzing articulation problems. EPG is simply an artificial palate put into the mouth of the client to detect the movement of tongue during speech production [3]. The major drawback of this device is that the client is not able to produce the speech sounds in a natural manner during the speech training session. The spectrograph is used by the SLP to analyze the speech production pattern of client. However, its time-frequency analysis is not user-friendly to the client because the analyzing result is difficult to be interpreted by the client. Realizing the problems and limitations of the conventional articulation training and speech analyzing tools, thus we propose a computer-based articulation speech training system to be more user-friendly and easy for use. The proposed system eliminates the use of candles, feathers, cotton balls and mirrors in the training sessions. It allows the client undergoes the articulation training by just talking to the microphone, which is connected to the computer system. The system provides the audio and visual feedback to help the client to identify his articulation pattern. The system enables the training session to be conducted without the presence of SLP, thus erasing the SLP’s laborious and repetitious job of monitoring. The use of computer technology in speech therapy and assessment is still quite new in Malaysia. The current available computer-based speech therapy systems in the Clinic of Audiology and Speech Sciences, Universiti Kebangsaan Malaysia (UKM), Malaysia are the IBM Speech Viewer [4] and Kay Elemtrics VisiPitch. These systems are designed mainly for voice therapy and are not ICICS-PCM 2003 15-18 December 2003 Singapore 0-7803-8185-8/03/$17.00 © 2003 IEEE 3A4.3