Researchers at the University of Nottingham and Kingston University have developed a 3D reconstruction tool which can make any image into a 3D GIF. What makes this so astounding is that you would usually need a series of 2D images to create this 3D version, but by using something called Convolutional Neural Network (CNN), this can work by using just one 2D image.
The research team said that in making the tool, they found incredible difficulty and faced “a number of methodological challenges such as establishing dense correspondences across large facial poses, expressions, and non-uniform illumination.”.
We’ve tested it out using a photo of Rob Da Bank, and as you can see it really does create a true to life image of the subject. If you would like to try this tool using one of your own ‘selfies’ or an image of someone else, perhaps your favourite celebrity, then click here to upload a photo and get started. For best results it is recommended that you upload a photo of the face looking directly at the camera just as you would for a passport photo.