Half-Temporal and Half-Frequency Attention U2Net for Speech Signal Improvement
Zehua Zhang, Shiyun Xu, Xuyi Zhuang, Yukun Qian, Lianyu Zhou, Mingjiang Wang
Abstract
During communication, volume changes, noise, and reverberation can disturb speech signals, significantly affecting the quality and intelligibility of speech. In the context of the ICASSP 2023 Signal Processing Grand Challenge, the first Speech Signal Improvement Grand Challenge (SIG) is organized to improve the quality of speech signals during communication. This paper proposes half-temporal and half-frequency attention U <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> Net for improving full-band speech signal. Channel-spectrum attention is proposed for the skip connection between the encoder and decoder. The proposed model achieves 0.353, 1.289, 0.604, 0.625, and 0.924 improvements in signal, noise, overall, reverberation, and loudness, respectively, in the SIG subjective test. The proposed model achieved fourth place in the SIG real-time track, showing excellent denoising and de-reverberation performance.
BibTeX
@inproceedings{icassp2023_halftemporalandh,
title = {Half-Temporal and Half-Frequency Attention U2Net for Speech Signal Improvement},
author = {Zehua Zhang and Shiyun Xu and Xuyi Zhuang and Yukun Qian and Lianyu Zhou and Mingjiang Wang},
booktitle = {ICASSP 2023},
year = {2023}
}