Recognition of human speech using q-Bernstein polynomials

Abstract

Recognition of human speech has long been a hot topic among artificial intelligence and signal processing researchers. In this paper, Pattern recognition and q-Bernstein polynomials have been combined to create computer software designed for the recognition of human speech, and the main principles of speech recognition have been explained in a systematic content. qBernstein polynomials, which are mathematical operators, have been applied for pattern recognition, and a new method has thus been developed. Software has been prepared using the Delphi 7 programming language, and with this software, this method has been applied for the processing of verbal expression recognition. In the program as developed, 16 000 samples of 8-bit stereo images were computerized. Speech recognition tests were conducted for six words, and information related to the results of this test is provided in this paper.