arXiv Open Access 2022

A light-weight full-band speech enhancement model

Qinwen Hu Zhongshu Hou Xiaohuai Le Jing Lu

Lihat Sumber

Abstrak

Deep neural network based full-band speech enhancement systems face challenges of high demand of computational resources and imbalanced frequency distribution. In this paper, a light-weight full-band model is proposed with two dedicated strategies, i.e., a learnable spectral compression mapping for more effective high-band spectral information compression, and the utilization of the multi-head attention mechanism for more effective modeling of the global spectral pattern. Experiments validate the efficacy of the proposed strategies and show that the proposed model achieves competitive performance with only 0.89M parameters.

Topik & Kata Kunci

eess.AS cs.SD

Penulis (4)

Qinwen Hu

Zhongshu Hou

Xiaohuai Le

Jing Lu

Format Sitasi

APA MLA BibTeX

Hu, Q., Hou, Z., Le, X., Lu, J. (2022). A light-weight full-band speech enhancement model. https://arxiv.org/abs/2206.14524

Akses Cepat

Lihat di Sumber

Informasi Jurnal

Tahun Terbit: 2022
Bahasa: en
Sumber Database: arXiv
Akses: Open Access ✓