Speech enhancement using group complementary joint sparse representations in modulation domain

The internal group structure of signals has been considered for some speech enhancement (SE) algo-rithms, but most of them are conducted in acoustic domain. In this paper, we propose to incorporate the group structure in modulation domain as prior information for complementary joint sparse represen-tations (CJSR). The modulation transform is applied to generate a set of sub-band amplitude spectrums with different modulation frequencies, which contain the novel time-frequency (TF) distributions differ-ent from that in acoustic domain. For each of these spectrums, we learn a couple of joint dictionaries in which the atoms are clustered in groups. The resulted dictionaries have structured characteristics of speech and noise. To represent a signal, we use an objective function based on sparse group lasso to acti-vate atoms on group level. By doing so, the speech is robustly recovered from mixture according to preset group pattern. The results of ablation study show that each part of proposed method, that is, modulation -domain processing and group sparsity, has its benefits for CJSR and combining both parts leads to a fur-ther performance improvement. In the final comparative experiment, the results show that the proposed method produces better objective speech quality, improving PESQ by 6.0% and segSNR by 12.7% com-pared with baseline method.(c) 2022 Elsevier Ltd. All rights reserved.

Automated summary

This paper proposes a method to incorporate group structure into joint sparse representations in the modulation domain to enhance signals. Experimental results show that this method performs better in terms of speech quality and can improve PESQ and segSNR scores.