LSF mapping for voice conversion with very small training sets

Abstract

To make voice conversion usable in practical applications, the number of training sentences should be minimized. With traditional Gaussian mixture model (GMM) based techniques small training sets lead to over-fitting and estimation problems. We propose a new approach for mapping line spectral frequencies (LSFs) representing the vocal tract. The idea is based on inherent intra-frame correlations of LSFs. For each target LSF, a separate GMM is used and only the source and target LSF elements best correlating with the current LSF are used in training. The proposed method is evaluated both objectively and in listening tests, and it is shown that the method outperforms the conventional GMM approach especially with very small training sets.

Arabic speech transformation using MFCC in GMM

“…This process will increase the energy of signal at higher frequency [3] as shown in figure (2). ) ( * ) ( ) ( 2 n s a n s n s (1) where s(n) is the speech signal , s 2 (n) is the output signal and the value of a is usually between 0.9 and 1.0. The ztransform of the filter is …”

“…Examples of such features include MFCCs (Mel Frequency Cepstral Coefficients) and LSFs (Line Spectral Frequencies) [1]. The aim of this paper is to extract the MFCC, and use them by the GMM for voice conversion of Arabic spoken words.…”

On using non-linear canonical correlation analysis for voice conversion based on Gaussian mixture model

“…The Line Spectral Frequencies (LSF) were selected as the representation of the vocal characteristics of source and target speakers due to their favorable interpolation properties and stableness [16] . A 20 ms length Hanning window overlapped by 10 ms was used to compute and extract the LPC parameters.…”

Mixture of Factor Analyzers Using Priors From Non-Parallel Speech for Voice Conversion

“…The joint density Gaussian mixture model (JD-GMM) [4], [5], [6] is one of the most effective approaches. Unfortunately, it requires relatively large parallel training data to avoid over-fitting [8].…”