[Software] Adobe uses AI to separate voice from background sounds in recordings into individual tracks

[[X]pErT-]

Adobe is working on a technology based on artificial intelligence (AI) that allows you to separate background noise from the voice in the audio recording of a video and improve the most important parts of it.

 

Project Sound Lift is one of the tools that Adobe showed off as a preview this Wednesday at its MAX event in Japan. At the moment it is a prototype, which has shared the potential it has for audio processing in videos.

 

 

   In this sense, Project Sound Lift uses AI to separate an audio recording into different tracks of voices, non-voice sounds and other background noises, as the company explains in a press release . Each track canalso be managed independently to improve the most important parts

 

   The novelty of this tool is that it works with different types of real-world environments, where there are a multitude of superimposed sounds and cannot always be controlled, and not only with clean audio inputs, as is the case, according to Adobe, with previous AI models. .

 

https://www.europapress.es/portaltic/software/noticia-adobe-usa-ia-separar-pistas-individuales-voz-sonidos-fondo-grabaciones-20231116141023.html