Learning to rank music tracks using triplet loss

05/18/2020
by   Laure Prétet, et al.
0

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.

READ FULL TEXT
research
10/29/2020

Learning Audio Embeddings with User Listening Data for Content-based Music Recommendation

Personalized recommendation on new track releases has always been a chal...
research
02/04/2022

Musical Audio Similarity with Self-supervised Convolutional Neural Networks

We have built a music similarity search engine that lets video producers...
research
08/11/2020

Content-based Music Similarity with Triplet Networks

We explore the feasibility of using triplet neural networks to embed son...
research
10/22/2020

Mood Classification Using Listening Data

The mood of a song is a highly relevant feature for exploration and reco...
research
08/10/2019

Personalized Music Recommendation with Triplet Network

Since many online music services emerged in recent years so that effecti...
research
06/07/2017

The Effects of Noisy Labels on Deep Convolutional Neural Networks for Music Tagging

Deep neural networks (DNN) have been successfully applied to music class...
research
09/04/2017

Musical NeuroPicks: a consumer-grade BCI for on-demand music streaming services

We investigated the possibility of using a machine-learning scheme in co...

Please sign up or login with your details

Forgot password? Click here to reset