Adobe introduced a new platform for generating and editing audio content
[ad_1]
American company Adobe announced Project Music GenAI Control AI tool, which is capable of both generating audio based on a text request and editing existing tracks. The project is under development, with scientists from the University of California and Carnegie Mellon participating in its creation, while the tool does not even have a user interface.
According to the developers, Project Music GenAI Control will turn the user into a conductor of a virtual orchestra and will help bring their ideas to life even for those who are not composers. Thanks to AI, tools will be available to regulate the structure of music tracks, tempo, as well as the intensity and repetition of patterns. In theory, this will allow you to create compositions with endless repetitions or remixes.
The company trains AI based on licensed or public domain data, and Adobe is also working on a watermark system that will make it easy to identify that content was created by AI.
The idea of creating AI that generates music based on a text request is not new: in December 2023, similar functionality from Microsoft added in the AI assistant Copilot. In July 2023, Meta (recognized as extremist in the Russian Federation and banned) presented similar AI MusicGen, and in May 2023 Google announced music generator MusicLM.
About how developers teach artificial intelligence to predict human feelings – in the publication “Kommersant” “AI in joy, AI in sorrow”.
[ad_2]
Source link