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

facebooktwittermailshare

ONLINE SINGING VOICE SEPARATION USING A RECURRENT ONE-DIMENSIONAL U-NET TRAINED WITH DEEP FEATURE LOSSES

Abstract: 

This paper proposes an online approach to the singing voice separation problem. Based on a combination of one-dimensional convolutional layers along the frequency axis and recurrent layers to enforce temporal coherency, state-of-the-art performance is achieved. The concept of using deep features in the loss function to guide training and improve the model’s performance is also investigated.

up
0 users have voted:

Paper Details

Authors:
Clement S. J. Doire
Submitted On:
9 May 2019 - 3:05am
Short Link:
Type:
Poster
Event:
Presenter's Name:
Clement S. J. Doire
Paper Code:
MLSP-P15.8
Document Year:
2019
Cite

Document Files

Poster presentation OR-U-Net

(40)

Subscribe

[1] Clement S. J. Doire, "ONLINE SINGING VOICE SEPARATION USING A RECURRENT ONE-DIMENSIONAL U-NET TRAINED WITH DEEP FEATURE LOSSES", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4160. Accessed: Sep. 16, 2019.
@article{4160-19,
url = {http://sigport.org/4160},
author = {Clement S. J. Doire },
publisher = {IEEE SigPort},
title = {ONLINE SINGING VOICE SEPARATION USING A RECURRENT ONE-DIMENSIONAL U-NET TRAINED WITH DEEP FEATURE LOSSES},
year = {2019} }
TY - EJOUR
T1 - ONLINE SINGING VOICE SEPARATION USING A RECURRENT ONE-DIMENSIONAL U-NET TRAINED WITH DEEP FEATURE LOSSES
AU - Clement S. J. Doire
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4160
ER -
Clement S. J. Doire. (2019). ONLINE SINGING VOICE SEPARATION USING A RECURRENT ONE-DIMENSIONAL U-NET TRAINED WITH DEEP FEATURE LOSSES. IEEE SigPort. http://sigport.org/4160
Clement S. J. Doire, 2019. ONLINE SINGING VOICE SEPARATION USING A RECURRENT ONE-DIMENSIONAL U-NET TRAINED WITH DEEP FEATURE LOSSES. Available at: http://sigport.org/4160.
Clement S. J. Doire. (2019). "ONLINE SINGING VOICE SEPARATION USING A RECURRENT ONE-DIMENSIONAL U-NET TRAINED WITH DEEP FEATURE LOSSES." Web.
1. Clement S. J. Doire. ONLINE SINGING VOICE SEPARATION USING A RECURRENT ONE-DIMENSIONAL U-NET TRAINED WITH DEEP FEATURE LOSSES [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4160