MaskCycleGAN-based Whisper to Normal Speech Conversion

TL;DR

This paper introduces a MaskCycleGAN-based method for converting whispered speech to normal speech, demonstrating improved performance through mask tuning and pre-processing, evaluated on the wTIMIT dataset with objective and subjective metrics.

Contribution

The study presents a novel application of MaskCycleGAN for whisper to normal speech conversion, highlighting the importance of mask parameter tuning and pre-processing for enhanced results.

Findings

MaskCycleGAN outperforms existing methods in speech conversion quality.

Tuning mask parameters improves conversion performance.

Pre-processing with voice activity detection enhances results.

Abstract

Whisper to normal speech conversion is an active area of research. Various architectures based on generative adversarial networks have been proposed in the recent past. Especially, recent study shows that MaskCycleGAN, which is a mask guided, and cyclic consistency keeping, generative adversarial network, performs really well for voice conversion from spectrogram representations. In the current work we present a MaskCycleGAN approach for the conversion of whispered speech to normal speech. We find that tuning the mask parameters, and pre-processing the signal with a voice activity detector provides superior performance when compared to the existing approach. The wTIMIT dataset is used for evaluation. Objective metrics such as PESQ and G-Loss are used to evaluate the converted speech, along with subjective evaluation using mean opinion score. The results show that the proposed approach offers considerable benefits.