Open AccessA Single Channel End-to-End Speech Enhancement using Complex Operations
Author(s) -
Jie Wu,
Hongqing Liu,
Gan Liu,
Yi Zhou
Publication year - 2022
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/2218/1/012001
Subject(s) - computer science , end to end principle , task (project management) , speech enhancement , channel (broadcasting) , domain (mathematical analysis) , time domain , encoder , speech recognition , decoding methods , algorithm , artificial intelligence , telecommunications , engineering , computer vision , mathematics , noise reduction , mathematical analysis , systems engineering , operating system
This paper investigates the possibility of using complex operations to perform speech enhancement task in time domain. To that end, first, the Hilbert transform is utilized to prepare the complex input in time domain. After that, the complex temporal convolutional network (CTCN) is developed to conduct complex convolutions. By cascading the TCN and the CTCN modules, the final proposed network form an encoder-decoder structure, which performs an end-to-end speech enhancement task. The results demonstrate that utilizing complex information in time domain indeed improves the enhancement performance. Compared to other approaches, the proposed network also demonstrates a superior performance in terms of objective evaluations.