Friday, 24 March, 2023

Meta builds three new AI models designed to make sound more realistic in VR experiences

<p>Three artificial intelligence (AI) models have been created by Meta to improve the audio realism of mixed and virtual reality experiences. The three AL models&mdash;Visual-Acoustic Matching, Visually-Informed Dereverberation, and VisualVoice&mdash;concentrate on human voice and sounds in video and are intended to move &quot;us toward a more immersive reality at a faster rate,&quot;. Sound in the Metaverse will be perceived differently depending on the acoustics.</p>