A company in the UK is trying to produce a computer that can visualize an item in the way that eyes’ can. If successful, it could be used to generate items for a virtual world. The company involved – Magic Pony Technology – has found a way to turn low-resolution videos into high- resolution ones. The computer can recognize the difference between the two once a high-resolution video has been fed into it. This is not an entirely new process, but now it can be done on an ordinary graphic processor and applications can be opened.
One of the founders of Magic Pony Rob Bishop has stated that there are some large companies interested in licensing the new technology. As video streaming, companies need video compression, one of the first things needed was tools to show how image quality can be improved by the use of deep learning and GPU’s that are fast and mobile. Rob went on to explain that smartphones could have better quality pictures taken on them especially when there is not much light.
There is no need for manually labeled images as statistical patterns are recognized and the difference in high or low resolution is noted, and it then determines how textures, edges or other features should appear.
The Paper will be presented later in 2016, and yet there have been significant improvements since it was written. In a final comment, the company name is explained- no one would believe in a “magic pony” unless they saw it and the technology is the same.
