Adobe Project Music GenAI Control – AI tool that generates audio from text descriptions

Adobe has unveiled Project Music GenAI Control, a platform capable of generating audio from text descriptions such as “happy dance” or “sad jazz” as well as a reference melody.

This tool allows users to customize results within a single workflow, including adjusting parameters such as tempo, intensity, repeating patterns and structure. Users can also extend a track to an arbitrary length, remix the music, or create endless loops.

Project Music GenAI Control was developed by Adobe in collaboration with researchers from the University of California and Carnegie Mellon University. This is currently a research project, but may become publicly available in the future. The platform does not yet have a user interface.

Adobe typically develops generative art tools on licensed or open-access data to avoid potential intellectual property issues. The company is also working on audio identification watermarking technology created by Project Music GenAI Control, but that process is still ongoing.