Downloads: 118 | Views: 196
Research Paper | Electronics & Communication Engineering | India | Volume 4 Issue 5, May 2015 | Rating: 6.7 / 10
Performance Evolution of Combine Approach of MFCC & DTW Technique for Speech Recognition
Nilesh Patel | D. G. Agrawal
Abstract: This paper proposes on speech recognition for English word using Mel Frequency Cepstral Coefficients (MFCCs) and Dynamic Time Warping (DTW) introduced by Sakoe Chiba. MFCC are the coefficients which collectively represent the short-term power spectrum of a sound. That Power spectrum based on a linear cosine transform of a log power spectrum on a nonlinear Mel scale of frequency. Start point, Short time Energy and Zero crossing detector and end point detection using Mel Frequency Cepstral Coefficients and Dynamic Time Warping algorithms used for recognition of samples English word which would be by different male and female speakers. The algorithm is test for sample speech signal which are recorded. This system is used for different application for like security system and also for controlling of robot. The result shows that accuracy is all different English word is above 61.66 % and also shows a comparison for different method applied for MFCC.
Keywords: STE, ZCR, MFCC, DTW
Edition: Volume 4 Issue 5, May 2015,
Pages: 1111 - 1114