This paper focuses on the idea of establishing a system that integrates the speech recognition, neural networks and facial modelling to benefit the human machine interaction in future. The utilization of Artificial Networks to enable communication between human and machines has been an intensive area of study of late. This area has been researched in different perspectives but still imposes strong challenge for researchers to find the most effective solution that allow real time communication between Man and Machine. It is imperative that this solution should be embedded with effective speech recognition and an interface for human machine interaction. It is also critical that this solution remains computationally less intensive for its applications as product in various fields. Thus this paper is proposed with all the criteria in place and with extensive analysis related to performance. This paper proposes a novel system equipped with an artificial neural network driven speech recognition, automatic lip synchronisation to create a realistic human model in the field of human machine interaction research. The system is expected to outperform the existing systems and is optimal on terms of complexity and the accuracy.
|Title of host publication
|Recent Advances in Intelligent Systems and Signal Processing
|World Scientific and Engineering Academy and Society
|Number of pages
|Published - 1 Dec 2003