Dynamic Extraction Method of 3D Parameters of Tongue for Pronunciation Recovery About Impaired Hearing Children

Abstract

3D talking head technology is applied in speech training of impaired hearing children, and dynamic extraction method of 3D parameters of tongue in 3D talking head is posed. 3D talking head visualization auxiliary teaching technology is a special model for autonomous language training without human intervention, the difficult point of which is the characteristics acquisition of the motion track of internal pronunciation organs (tongues and jaws)when pronouncing. According to the five different modes of tongue movement when pronouncing, the specific dynamic extraction methods of 3D tongue and jaw features corresponding to the five modes are posed and applied in the pronunciation recovery about impaired hearing children. The experimental results show that the collection efficiency of dynamic parameter and accuracy of characteristics acquisition is enhanced by this method and also this method has strong antijamming capability.