A challenging task in recognizing the speech of the hearing impaired using normal hearing models in classical Tamil language

This paper discusses the development of robust speech recognition system for recognizing the speech of the children with hearing impairment with no necessity to collect large amount of speech samples from them. they normally use sign language to communicate with others though they have perfect vocal structure to produce speech in the form of sequence of sounds. so,it is necessary to develop the system for recognizing their speeches especially in the native language and we have considered tamil speaking hearing impaired children in our work. it is normally a challenging task to recognize the distorted speeches of hearing impaired speakers. but here the performance of the system is analyzed by applying the hearing impaired speeches directly on the models developed using the speeches of the normal speakers. this work uses mel frequency perceptual linear predictive cepstrum as a feature at the front end and k means clustering at the back end for recognizing the speeches of the hearing impaired and the recognized speech uttered by normal speakers can be displayed as a visual output or audio output which can be clearly understandable by others. the approach to this issue employed rls filtering as an additional preprocessing technique and correlation analysis is done between the rls filtered speech and clean normal speech to select the corresponding clean normal speech from the set of several normal speeches. hearing impaired speeches will be heard and interpreted clearly by the normal speakers and the system can be used to help the hearing impaired to improve their status in society.