English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Speech bandwidth extension (BWE) refers to widening the frequency bandwidthrange of speech signals, enhancing the speech quality towards brighter andfuller. This paper proposes a generative adversarial network (GAN) based BWEmodel with parallel prediction of Amplitude and Phase spectra, named AP-BWE,which achieves both high-quality and efficient wideband speech waveformgeneration. The proposed AP-BWE generator is entirely based on convolutionalneural networks (CNNs). It features a dual-stream architecture with mutualinteraction, where the amplitude stream and the phase stream communicate witheach other and respectively extend the high-frequency components from the inputnarrowband amplitude and phase spectra. To improve the naturalness of theextended speech signals, we employ a multi-period discriminator at the waveformlevel and design a pair of multi-resolution amplitude and phase discriminatorsat the spectral level, respectively. Experimental results demonstrate that ourproposed AP-BWE achieves state-of-the-art performance in terms of speechquality for BWE tasks targeting sampling rates of both 16 kHz and 48 kHz. Interms of generation efficiency, due to the all-convolutional architecture andall-frame-level operations, the proposed AP-BWE can generate 48 kHz waveformsamples 292.3 times faster than real-time on a single RTX 4090 GPU and 18.1times faster than real-time on a single CPU. Notably, to our knowledge, AP-BWEis the first to achieve the direct extension of the high-frequency phasespectrum, which is beneficial for improving the effectiveness of existing BWEmethods.

Towards High-Quality and Efficient Speech Bandwidth Extension with Parallel Amplitude and Phase Prediction