Page 1 of 1

How to nearly automatic make a 3D-Modell from a persons picture.

Posted: Wed Jul 01, 2020 7:10 am
by Theo_Gottwald
Can we nearly automatic make a 3D-Modell from a persons 2D-Picture?
You may think "This is impossible".
But then you never realized that Companies like the CIA and Facebook for long time have Face- and Body recognition technologies that can recognize you perfectly from a 2D-Picture.
Sometimes these technologies leak ... and so.

Video on the topic

We introduce a multi-level framework that infers 3D geometry of clothed humans at an unprecedentedly high 1k image resolution in a pixel-aligned manner, retaining the details in the original inputs without any post-processing.

Here is how it works:
2D-Picture to 3D Modell using Neural networks

This is the Link to the Github:
2D to 3D Github

As its Python based someone could make a Blender AddOn with it,

There is also a Googled Version which can be used Online to some degree.
Google-Collab-Version

PS: Thinking about it there are many uses for this technology, even the police could use it to get a better view on Gangsters that have been photographed only from one direction.