Toward Universal Text-to-Music Retrieval
Authors:
- SeungHeon DohSDSeungHeon Doh
- Minz WonMWMinz Won
- Keunwoo ChoiKCKeunwoo Choi
- Juhan NamJNJuhan Nam
Preprint
Abstract
This paper introduces effective design choices for text-to-music retrieval systems. An ideal text-based retrieval system would support various input queries such as pre-defined tags, unseen tags, and sentence-level descriptions. In reality, most previous works mainly focused on a single query type (tag or sentence) which may not generalize to another input type. Hence, we review recent text-based music retrieval systems using our proposed benchmark in two main aspects: input text representation and training objectives. Our findings enable a universal text-to-music retrieval system that achieves comparable retrieval performances in both tag- and sentence-level inputs. Furthermore, the proposed multimodal representation generalizes to 9 different downstream music classification tasks. We present the code and demo online.Keywords
All Related Versions
This publication has 4 references indexed in Scilit:
- Deep Learning for Audio-Based Music Classification and Tagging: Teaching Computers to Distinguish Rock from BachIEEE Signal Processing Magazine, 2018
- Studying emotion induced by music through a crowdsourcing gameInformation Processing & Management, 2016
- Lyrics-Based Audio Retrieval and Multimodal Navigation in Music CollectionsLecture Notes in Computer Science, 2007
- Musical genre classification of audio signalsIEEE Transactions on Speech and Audio Processing, 2002