Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

Using recurrences in time and frequency within U-net architecture for speech enhancement

Abstract: 

When designing fully-convolutional neural network, there is a trade-off between receptive field size, number of parameters and spatial resolution of features in deeper layers of the network. In this work we present a novel network design based on combination of many convolutional and recurrent layers that solves these dilemmas. We compare our solution with U-nets based models known from the literature and other baseline models on speech enhancement task. We test our solution on TIMIT speech utterances combined with noise segments extracted from NOISEX-92 database and show clear advantage of proposed solution in terms of SDR (signal-to-distortion ratio), SIR (signal-to-interference ratio) and STOI (spectro-temporal objective intelligibility) metrics compared to the current state-of-the-art.

up
0 users have voted:

Paper Details

Authors:
Szymon Drgas
Submitted On:
8 May 2019 - 9:13am
Short Link:
Type:
Poster
Event:
Presenter's Name:
Tomasz Grzywalski
Paper Code:
3235
Document Year:
2019
Cite

Document Files

Grzywalski_Drgas.pdf

(52)

Subscribe

[1] Szymon Drgas, "Using recurrences in time and frequency within U-net architecture for speech enhancement", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4091. Accessed: Dec. 08, 2019.
@article{4091-19,
url = {http://sigport.org/4091},
author = {Szymon Drgas },
publisher = {IEEE SigPort},
title = {Using recurrences in time and frequency within U-net architecture for speech enhancement},
year = {2019} }
TY - EJOUR
T1 - Using recurrences in time and frequency within U-net architecture for speech enhancement
AU - Szymon Drgas
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4091
ER -
Szymon Drgas. (2019). Using recurrences in time and frequency within U-net architecture for speech enhancement. IEEE SigPort. http://sigport.org/4091
Szymon Drgas, 2019. Using recurrences in time and frequency within U-net architecture for speech enhancement. Available at: http://sigport.org/4091.
Szymon Drgas. (2019). "Using recurrences in time and frequency within U-net architecture for speech enhancement." Web.
1. Szymon Drgas. Using recurrences in time and frequency within U-net architecture for speech enhancement [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4091