In this paper, we propose a novel neutral-to-emotional voice conversion (VC) model that can effectively learn a mapping from neutral to emotional speech with limited emotional voice data. Although conventional VC techniques have achieved tremendous success in spectral conversion, the lack of representations in fundamental frequency (F0), which explicitly represents prosody information, is still a major limiting factor for emotional VC. To overcome this limitation, in our proposed model, we outline the practical elements of the cross-wavelet transform (XWT) method, highlighting how such a method is applied in synthesizing diverse representations of F0 features in emotional VC. The idea is (1) to decompose F0 into different temporal level representations using continuous wavelet transform (CWT); (2) to use XWT to combine different CWT-F0 features to synthesize interaction XWT-F0 features; (3) and then use both the CWT-F0 and corresponding XWT-F0 features to train the emotional VC model. Moreover, to better measure similarities between the converted and real F0 features, we applied a VA-GAN training model, which combines a variational autoencoder (VAE) with a generative adversarial network (GAN). In the VA-GAN model, VAE learns the latent representations of high-dimensional features (CWT-F0, XWT-F0), while the discriminator of the GAN can use the learned feature representations as a basis for a VAE reconstruction objective.

This paper is concerned with the problem of determining the typical features of a curve when it is observed with noise. It has been shown that one can characterize the Lipschitz singularities of a signal by following the propagation across scales of the modulus maxima of its continuous wavelet transform. A nonparametric approach, based on appropriate thresholding of the empirical wavelet coefficients, is proposed to estimate the wavelet maxima of a signal observed with noise at various scales. In order to identify the singularities of the unknown signal, we introduce a new tool, “the structural intensity”, that computes the “density” of the location of the modulus maxima of a wavelet representation along various scales. This approach is shown to be an effective technique for detecting the significant singularities of a signal corrupted by noise and for removing spurious estimates. The asymptotic properties of the resulting estimators are studied and illustrated by simulations. An application to a real data set is also proposed.