Wouldn’t it be nice to not spend hours editing your pictures in apps like Lightroom and Photoshop before uploading them on Facebook, Instagram or any other social network? Well, soon you won’t have to.
A new automatic photo retouching app has been created by MIT researchers that can improve your images in real-time. It does its work even before you click on that camera of yours!
This new app created by MIT’s Computer Science and Artificial Intelligence Laboratory and Google is so energy-efficient that you can run it on your smart phone and correct and retouch the picture at run-time without having to go through the painstaking post-shoot editing.
This image processing system uses deep learning to determine the best way to retouch images, using before and after examples. It basically corrects and retouches images as a casual photographer when you are still framing the shot.
A system takes care of the app’s deep learning capabilities. It goes through thousands of training data to learn, analyzing different raw and retouched photographs. For the training data, the researchers created 5000 corrected and retouched images for the training set with Adobe Systems software.
A model is generated based upon these training samples, making the neural network learn what kind of end result is expected, given a particular kind of raw image.
Michaël Gharbi, an MIT graduate and researcher, said:
Google heard about the work I’d done on the transform recipe. They themselves did a follow-up on that, so we met and merged the two approaches.
The idea was to do everything we were doing before but, instead of having to process everything on the cloud, to learn it. And the first goal of learning it was to speed it up.
Nothing has been announced about the release of this new app as of yet. It would be nice if the technology gets licensed out to other camera apps allowing for retouching of photos in real-time.