Abstract
We propose the development of an innovative AI tool specifically designed to transform Jamaican music and other genres into clean radio-edit versions suitable for airplay. Utilizing advanced generative AI technology, our tool aims to sanitize songs while retaining their intrinsic essence. In this paper, we leverage modern generative AI techniques to recreate airplay-friendly versions within the same musical genre, focusing on Jamaican music's unique cultural significance. Our evaluation, conducted across multiple songs, reveals that the tool demonstrates commendable efficacy in its transformative task. However, ongoing refinements are necessary to ensure the tool consistently maintains the distinctive style and mood of the original songs.
Recommended Citation
Stone, Matthew and Mansingh, Gunjan, "AI tool for cleaning up lewd music: A Jamaican Perspective (Research- in-Progress)" (2024). CONF-IRM 2024 Proceedings. 26.
https://aisel.aisnet.org/confirm2024/26