当前位置:首页 >> >>

A Speech Orientation Animation System to Disabled People


A Speech Orientation Animation System to Disabled People
JESULIANA NASCIMENTO ULYSSES, AURA CONCI Instituto de Computa??o – UFF, R Passo da Pátria 156, 24210-240, Niterói, RJ, Brasil nul

ysses@artnet.com.br , aconci@ic.uff.br Abstract. Audio disabled children or adults under oral rehabilitating therapy learn how to treat speech disorders or even how utter sounds imitating mouth elements position and movement. This technique has been proved beneficial to help people since specialized instructional program adopt it. The main proposal of this work is the development of an affordable tool to aid in this therapy. 1 Introduction A technique to teach deaf people is by lips reading. They can know how to speech by emitting sounds at the same time that they mimic mouth elements and movement. The therapy constitutes of in vivo teaching and auto-training sessions, and must rely on previous researches on what is the influence of each mouth element on the formation of the basic sound of a language (Warren and Kaiser, 1988). 2 Development The controlled facial animation is a complex task due to the huge complexity of the real structure of the face that is composed of muscle, bone and skin. The facial expression is related with speech mainly by mouth shape. The muscle movements control the lips and other elements of the face. The possible movements are caused by contraction or relax of muscles linear and sphincter. A linear muscle becomes thickened and shortened with contractions, and pull the face in a direction. Sphincter muscles are circular band muscles that encircle and close an orifice of the body or one of its hollow organs. On the face they surround the mouth closing the lips. Complex expressions are due to combination of these unidirectional movements. This work uses the Waters' abstract musclebased model (http://www.crl.research.digital.com/ projects/facial/facial.html) to control the synthetic face movement. That is, the linear and sphincter muscles can adequately control the mouth. In addition to these two types muscle, a control for the jaw rotation was developed. Through these controls, muscles and jaw rotation, the polygon mesh can be deformed and thus produces large range labial expression. The implementation uses macros to create the animation of the mouth. These macros control duration of the expression using the influence zone of the muscle movements. These values attributed adequately to muscle control generate the phonetic animation. We developed a proper approach to measure the mouth position during the speech process. Four parameters were used to measure the mouth element position and shape: JAW, FLARE, CORNERS and EDGES (Rodman, et al., 1997). JAW is the jaw position and is measured as the distance between the two (upper and lower) teeth. FLARE is the height of the maximum vertical aperture between lips. The other two parameters characterize the horizontal movement of lips. The EDGE measures the lips joint points, where the upper and lower lips connect. CORNERS represent the horizontal aperture of the lips during speech. To implementation the system was used C++ using OpenGL, in PCs. The main objects of the face are: Face object, FaceIO object, FaceStructure object, Skin object, Expression object, ListeExpressions object and muscle hierarchy. 3 Results and Conclusion Although, we are at the initial steps in direction of a complete system, we can animate all vowel sounds and see that the presented methodology and the facial model used enable the construction of such system. Next steps in this direction will be systematized of the methodology of measurements for easy map of the other Portuguese phonetic sounds. Acknowledgment Our deepest gratitude goes to Emmanuel Tanguy and Fabio Policarpo for they help with the facial model and C++ codes of this model. Reference [1] S. Warren and A. Kaiser. Incidental language teaching. A critical review. Journal of Speech and Hearing Disorders 51 (1988), 291-299. [2] Rodman, D. McAllister, and D. Bitzer. Lip Synchronization as an Aid to the Hearing Disabled. Proceedings of the American Voice Input/Output Society 1997, 233 - 248.

Proceedings of the XV Brazilian Symposium on Computer Graphics and Image Processing (SIBGRAPI’02) 1530-1834/02 $17.00 ? 2002 IEEE


相关文章:
更多相关标签: