Automatic melody harmonization with triad chords

A comparative study

Bibliographic Data

ID14872276
AuthorsYin-Cheng Yeh (Academia Sinica, Taipei, Taiwan, corresponding author), Wen-Yi Hsiao (Academia Sinica, Taipei, Taiwan), Satoru Fukayama (0000-0001-6506-2796, National Institute of Advanced Industrial Science and Technology), Tetsuro Kitahara (Nihon University, Chiyoda, Japan), Benjamin Genchel (Georgia Institute of Technology), Hao-Min Liu (Academia Sinica, Taipei, Taiwan), Hao-Wen Dong (0000-0002-1020-0001, Academia Sinica, Taipei, Taiwan), Yian Chen (0000-0002-5971-9111, KKBOX Inc., Taipei, Taiwan), Yi‐An Chen (0000-0002-8002-6256), Terence Leong (KKBOX Inc., Taipei, Taiwan), Yi-Hsuan Yang (0000-0002-2724-6161, Academia Sinica, Taipei, Taiwan)
Year2021
Volume50
Issue1
Pages37-51
Publication date2021-01-01
Open AccessNo
TypeARTICLE
VenueJournal of New Music Research (JOURNAL)
Journal identifiersISSN: 0929-8215 • E-ISSN: 1744-5027
PublisherInforma UK Limited (PUBLISHER • GB)
DOI10.1080/09298215.2021.1873392
OpenAlexW3124507227
LanguageEN
References cited26

The task of automatic melody harmonization aims to build a model that generates a chord sequence as the harmonic accompaniment of a given multiple-bar melody sequence. In this paper, we present a comparative study evaluating the performance of canonical approaches to this task, including template matching, hidden Markov model, genetic algorithm and deep learning. The evaluation is conducted on a dataset of 9226 melody/chord pairs, considering 48 different triad chords. We report the result of an objective evaluation using six different metrics and a subjective study with 202 participants, showing that a deep learning method performs the best

Harmonization · Hidden Markov model · Machine learning · Natural language processing · Sequence labeling · Speech recognition · Computer Science · Music and Audio Processing · Music Technology and Sound Studies · Neuroscience and Music Perception · Artificial Intelligence

Tools

Open DOI