Disney Develops Machine Learning Tool to Create Realistic 3D Faces – Digital Information World

3D animation has become a cornerstone of the film industry because of the fact that this is the sort of thing that could potentially end up allowing studios to create movies without having to actually film them but instead develop them on various computers using a wide range of animation techniques. With all of that having been said and now out of the way, it is important to note that a lot of the 3D faces that you see are probably going to make you feel rather uncomfortably due to something called the uncanny valley.

This is a phenomenon wherein if something looks very similar to us without being entirely accurate, it would bring feelings of terror or discomfort when we see them. As a result of the fact that this is the case, a lot of animation studios have struggled to create 3D faces that are as realistic as possible. This is why so few 3D movies feature human beings that look realistic rather than cartoonish, although recent developments made by Disney might just make this sort of thing less of an issue all in all.

The machine learning tool would do this by better analyzing the various intricacies of the human face. There are a lot of subtle variations in our expressions, and another thing to note is that its very unlikely that two different people would end up having the exact same expressions as each other. Hence, the machine learning tool will help to broaden the available options that animators can currently take advantage of.

This shows just how relevant the world of tech is regardless of what industry you may be talking about at any given point in time. Disneys research into machine learning is primarily to help the company make movies but it will also have a widespread impact on other areas as well.

Read next: AI Could Soon Give Speakers Directional Voice Detection

Read the original here:
Disney Develops Machine Learning Tool to Create Realistic 3D Faces - Digital Information World

Related Posts
This entry was posted in $1$s. Bookmark the permalink.