Munehiro Namba

Learn More
This paper presents a novel algorithm that modifies the speech uttered by a source speaker to sound as if produced by a target speaker. In particular, we address the issue of transformation of the vocal tract characteristics from one speaker to another. The approach is based on estimating spectral envelopes using radial basis function (RBF) networks, which(More)
In this paper, we propose a novel method for detecting the fundamental frequencies of speech signals contaminated by noise. The proposed method exploits an eigen-based subspace principle to estimate unknown parameters of the noisy speech signal. In the proposed method, the estimated parameters are used for recovering the spectrum of the signal buried in(More)
In this paper, a wavelet transform domain realization of the blind equalization technique termed as EVA is applied to speech analysis. The conventional linear prediction problem can be viewed as a constrained blind equalization problem. Because the EVA does not impose any restriction to the probability distribution in the input (the glottal excitation), the(More)
The classical learning technique such as the back-propagation algorithm minimizes the expectation of the squared error that arise between the actual output and the desired output of supervised neural networks. The network trained by such a technique, however, does not behave in the desired way, when it is embedded in the system that deals with non-Gaussian(More)
  • 1