Attention-Guided Generative Adversarial Network for Whisper to Normal Speech Conversion
This work addresses speech quality and intelligibility issues for whispered speech conversion, which is incremental as it builds on existing methods with specific improvements.
The paper tackled converting whispered speech to normal speech by proposing an attention-guided generative adversarial network (AGAN-W2SC) that avoids fundamental frequency estimation and feature alignment, resulting in improved speech quality and intelligibility compared to dynamic-time-warping-based methods.
Whispered speech is a special way of pronunciation without using vocal cord vibration. A whispered speech does not contain a fundamental frequency, and its energy is about 20dB lower than that of a normal speech. Converting a whispered speech into a normal speech can improve speech quality and intelligibility. In this paper, a novel attention-guided generative adversarial network model incorporating an autoencoder, a Siamese neural network, and an identity mapping loss function for whisper to normal speech conversion (AGAN-W2SC) is proposed. The proposed method avoids the challenge of estimating the fundamental frequency of the normal voiced speech converted from a whispered speech. Specifically, the proposed model is more amendable to practical applications because it does not need to align speech features for training. Experimental results demonstrate that the proposed AGAN-W2SC can obtain improved speech quality and intelligibility compared with dynamic-time-warping-based methods.