Proceedings of the 2018 Conference of the North American Chapter Of the Association for Computational Linguistics: Hu 2018
DOI: 10.18653/v1/n18-1015
|View full text |Cite
|
Sign up to set email alerts
|

A Melody-Conditioned Lyrics Language Model

Abstract: This paper presents a novel, data-driven language model that produces entire lyrics for a given input melody. Previously proposed models for lyrics generation suffer from the inability of capturing the relationship between lyrics and melody partly due to the unavailability of lyrics-melody aligned data. In this study, we first propose a new practical method for creating a large collection of lyrics-melody aligned data and then create a collection of 1,000 lyrics-melody pairs augmented with precise syllable-not… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

0
38
1

Year Published

2019
2019
2023
2023

Publication Types

Select...
5
2
2

Relationship

1
8

Authors

Journals

citations
Cited by 41 publications
(39 citation statements)
references
References 10 publications
0
38
1
Order By: Relevance
“…However, the quality of the lyrics transcreated by the users should be evaluated. Watanabe et al [30] performed a qualitative evaluation that used a crowdsourcing mechanism and quantitative evaluation involving a test set of perplexity and a line/block boundary replication task for measuring the consistency between the melody and boundaries in the generated lyrics. Dzogang et al [31] focused on the task of detecting emotion in texts and analyzing texts as point sets in an emotional multidimensional space.…”
Section: Discussionmentioning
confidence: 99%
“…However, the quality of the lyrics transcreated by the users should be evaluated. Watanabe et al [30] performed a qualitative evaluation that used a crowdsourcing mechanism and quantitative evaluation involving a test set of perplexity and a line/block boundary replication task for measuring the consistency between the melody and boundaries in the generated lyrics. Dzogang et al [31] focused on the task of detecting emotion in texts and analyzing texts as point sets in an emotional multidimensional space.…”
Section: Discussionmentioning
confidence: 99%
“…• Overall (Watanabe et al 2018) What is the overall score of the melody? Results Table 3 shows the human evaluation results.…”
Section: Human Evaluationmentioning
confidence: 99%
“…Later, in Tra-la-Lyrics 2.0 [6], the words in the generated lyrics have higher semantic associations among them and with given seeds, while the rhythms are still matched if rhymes are present. Watanabe et al [7] used a melody-conditioned model to generate fluent lyrics while maintaining the compatibility between the boundaries of lyrics and melody structures.…”
Section: Introductionmentioning
confidence: 99%