The team at Google Brain has come up with a way to take low resolution pictures and turn them into images that look like high resolution photos. Now I know what you are asking. Google has a division called Google brain? I know, that must be the coolest job in the world.
The team uses a neural network that gathers countless images, and approximates what the low-resolution image should look like based on the high resolutions it is using as a comparison. They don’t always get it right, but they are in the early stages. What does always happen is the resolution of the images is vastly improved without the normal blurred edges that accompany previous attempts to enhance a photos image resolution.
This can prove to be very important for people with visual impairments who need to enhance an image without causing distortion. The full paper on the project from Google Brain can be found here.
