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

facebooktwittermailshare

Learning to rank music tracks using triplet loss

Abstract: 

Most music streaming services rely on automatic recommendation algorithms to exploit their large music catalogs. These algorithms aim at retrieving a ranked list of music tracks based on their similarity with a target music track. In this work, we propose a method for direct recommendation based on the audio content without explicitly tagging the music tracks. To that aim, we propose several strategies to perform triplet mining from ranked lists. We train a Convolutional Neural Network to learn the similarity via triplet loss. These different strategies are compared and validated on a large-scale experiment against an auto-tagging based approach. The results obtained highlight the efficiency of our system, especially when associated with an Auto-pooling layer.

up
0 users have voted:

Paper Details

Authors:
Laure Pretet, Gael Richard, Geoffroy Peeters
Submitted On:
18 May 2020 - 3:27am
Short Link:
Type:
Presentation Slides
Event:
Presenter's Name:
Laure Prétet
Paper Code:
AUD-P5.4
Document Year:
2020
Cite

Document Files

icassp.pdf

(49)

Subscribe

[1] Laure Pretet, Gael Richard, Geoffroy Peeters, "Learning to rank music tracks using triplet loss", IEEE SigPort, 2020. [Online]. Available: http://sigport.org/5389. Accessed: Jul. 14, 2020.
@article{5389-20,
url = {http://sigport.org/5389},
author = {Laure Pretet; Gael Richard; Geoffroy Peeters },
publisher = {IEEE SigPort},
title = {Learning to rank music tracks using triplet loss},
year = {2020} }
TY - EJOUR
T1 - Learning to rank music tracks using triplet loss
AU - Laure Pretet; Gael Richard; Geoffroy Peeters
PY - 2020
PB - IEEE SigPort
UR - http://sigport.org/5389
ER -
Laure Pretet, Gael Richard, Geoffroy Peeters. (2020). Learning to rank music tracks using triplet loss. IEEE SigPort. http://sigport.org/5389
Laure Pretet, Gael Richard, Geoffroy Peeters, 2020. Learning to rank music tracks using triplet loss. Available at: http://sigport.org/5389.
Laure Pretet, Gael Richard, Geoffroy Peeters. (2020). "Learning to rank music tracks using triplet loss." Web.
1. Laure Pretet, Gael Richard, Geoffroy Peeters. Learning to rank music tracks using triplet loss [Internet]. IEEE SigPort; 2020. Available from : http://sigport.org/5389