Home > Business > Melody-Conditioned Lyrics Generation with SeqGANs

Melody-Conditioned Lyrics Generation with SeqGANs

wallpapers Business 2020-11-24 >

The automatic lyrics generation from a given piece of music can provide musicians with inspirations. Current deep learning methods for this task only incorporate limited music information and cannot generate lyrics that fit given melodies.

A recent paper tackles this problem by learning the relation between lyrics and melody. Then, lyrics are generated from the input melody using sequence generative adversarial networks. The approach does not require pre-processing of data or prior knowledge.

Image credit: 刘睿忱 via Wikimedia (CC BY-SA 4.0)

Also, an extension that enables us to choose the theme of generated lyrics was created. It is shown that these additional restrictions have no negative impact on lyrics. The quality of generated lyrics outperforms a baseline. The lyrics most often are readable and work well with the melody. However, further improvements in rhyme and syllabic structure could be made.

Automatic lyrics generation has received attention from both music and AI communities for years. Early rule-based approaches have~—due to increases in computational power and evolution in data-driven models—~mostly been replaced with deep-learning-based systems. Many existing approaches, however, either rely heavily on prior knowledge in music and lyrics writing or oversimplify the task by largely discarding melodic information and its relationship with the text. We propose an end-to-end melody-conditioned lyrics generation system based on Sequence Generative Adversarial Networks (SeqGAN), which generates a line of lyrics given the corresponding melody as the input. Furthermore, we investigate the performance of the generator with an additional input condition: the theme or overarching topic of the lyrics to be generated. We show that the input conditions have no negative impact on the evaluation metrics while enabling the network to produce more meaningful results.

MIS-ASIA is an online content marketing platform that has a large number of visitors worldwide. It is considered to be the leading IT, mechanical, chemical, and nanomaterial information distributor in the Asia-Pacific region. The MIS-ASIA website provides high-quality articles and news on digital information technology, mechanical technology, nanotechnology, biology and science for scientists, engineers and industry experts, machinery suppliers and buyers, chemical suppliers and laboratories. If you need advertising and posting service, or you need to start sponsorship, please contact us.

Say something
  • All comments(0)
    No comment yet. Please say something!