Static Images Can Be Used To Create 3D Images By This Special Algorithm

Chung-Yi Weng, a Ph.D. student from the University of Washington, along with his peers have created an astounding software which could convert 2D images to 3D images.

The software known as “Photo Wake-Up” enables creating movements for a character object in a static picture frame without leaving a blank hole on the picture itself.

The developers used the method of identifying the main 2D object as the input, then creating the animated variation of the object. The created-animation can then perform basic movement like “sit, walk-out, jump, run” in 3D.

How we interact with 2D images could be redefined by this method. Also, the method offers a solution for forming a virtual avatar from a simple 2D-picture, as explaining by Weng and his colleague.

As we can see in the following video on YouTube, which shows the effect of the software. It can make a simple cartoon character jump up and down, as well as help Stephen Curry run out of the picture frame. These outputs prove that the software works amazingly. Here is the video:

The only requirement Weng's Photo Wake-up software needs to meet is a frame displaying a silhouette that wraps around the 3D shape, and as a result, the software runs considerably smoother.

Many researchers have devoted into creating similar effects prior, but no results have been as impressive as what Weng has done. A new ability can be developed from Weng’s approach, that is identifying body parts like arms and legs, and individually wrap them in a form that similarly coordinates with the 2D image.

The software can also be implemented in AR (augmented reality) and VR (virtual reality), both the technologies could one day reinvent our interaction with 2D photos or work of art the next time we view our family albums, or visit an art gallery.