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

In speech enhancement, complex neural network has shown promising performancedue to their effectiveness in processing complex-valued spectrum. Most of therecent speech enhancement approaches mainly focus on wide-band signal with asampling rate of 16K Hz. However, research on super wide band (e.g., 32K Hz) oreven full-band (48K) denoising is still lacked due to the difficulty ofmodeling more frequency bands and particularly high frequency components. Inthis paper, we extend our previous deep complex convolution recurrent neuralnetwork (DCCRN) substantially to a super wide band version -- S-DCCRN, toperform speech denoising on speech of 32K Hz sampling rate. We first employ acascaded sub-band and full-band processing module, which consists of twosmall-footprint DCCRNs -- one operates on sub-band signal and one operates onfull-band signal, aiming at benefiting from both local and global frequencyinformation. Moreover, instead of simply adopting the STFT feature as input, weuse a complex feature encoder trained in an end-to-end manner to refine theinformation of different frequency bands. We also use a complex feature decoderto revert the feature to time-frequency domain. Finally, a learnable spectrumcompression method is adopted to adjust the energy of different frequencybands, which is beneficial for neural network learning. The proposed model,S-DCCRN, has surpassed PercepNet as well as several competitive models andachieves state-of-the-art performance in terms of speech quality andintelligibility. Ablation studies also demonstrate the effectiveness ofdifferent contributions.

S-DCCRN: Super Wide Band DCCRN with learnable complex feature for speech enhancement