Title: Discrimination algorithm using voiced detection method and time-delay neural network system by 3 FFT sub-bands

Authors: Jae Seung Choi

Addresses: Department of Electronic Engineering, College of Engineering, Silla University, 140 Baegyang-daero Blvd., 700 Beon-gil Rd., Sasang-gu, Busan, 617-736, Korea

Abstract: In the field of speech processing, the main application of a neural network is the category discrimination of phoneme recognition and speech/speaker recognition. In a speech signal, information on the time variation is significant when training the neural network for the speech signal input. Therefore, this paper proposes a speech discrimination algorithm in noisy speech signals using a voiced detection method and a time-delay neural network with a time structure. Thus, this algorithm first detects voiced sections using the proposed neural network at each frame in the condition of background noises, then discriminates the speech signals using the time-delay neural network based on three fast Fourier transform sub-bands, in the noisy environments. The effectiveness of the proposed algorithm is experimentally confirmed based on measuring the correct discrimination rates for speech degraded by various noises.

Keywords: discrimination rate; voiced detection; speech discrimination; time-delay neural networks; TDNN; fast Fourier transform; FFT; background noise; speech processing; speech recognition; phoneme recognition; speaker recognition.

DOI: 10.1504/IJCVR.2015.068795

International Journal of Computational Vision and Robotics, 2015 Vol.5 No.2, pp.99 - 111

Received: 09 Jul 2013
Accepted: 13 Jan 2014
Published online: 13 Apr 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article