Architectural Approaches for Phonemes Recognition Systems

Abstract

Based on the literature, it is possible to build voice recognition systems by using voice synthesizers and voice command controls. In addition, phonemes recognition can be made by implementing algorithms already created for this kinds of tasks. Nevertheless, phonemes recognition might generate some errors, when the implementation of such algorithms is unsuitable. Then, the possibility to perform phonemes recognition based on open source APIs arises. In the work presented in this paper, we used open source APIs for voice commands recognition. Thus, we propose an architecture that allows the construction of a system for phonemes recognition and voice synthesizers. The results have been implemented and validated in order to illustrate the reliability of the proposed architecture.