AI tool enhances low-res pixelated images into hi-res

You see it in cop shows like CSI all the time. The police are trying to enhance pixelated images to read what seems like non-existent details in photos. As far-fetched as it may seem, researchers have been artificial intelligence technologies to help fill in missing image details.

One consumer-ready websites is already making some of this magic accessible. Let’s Enhance is a web-based software that creates higher-resolution files from low-res images without the normal JPEG artifacts. Using neural networks, this software can take pixelated images and increase the resolution with only minor quality loss. When compared to a traditionally upscaled image in Photoshop, the quality difference is impressive.

Let's Enhance: Low-res pixelated images into hi-res quality

Let’s Enhance works by first identifying the type of image, such as a portrait or landscape. It then uses a separate network to enhance the image using some of the anomalies of the image type. Along with using different networks for different image types, the program uses 3 different layers to produce the final image.

The first layer is the Anti-JPEG filter, dedicated to just removing the artifacts that are created. The second, called the Boring filter, does the actual upscaling while also working to preserve the details in the original pixelated images. The third layer is the one that makes the pixelated images look like they were shot at that original resolution in the first place—aptly called the Magic filter. This Magic filter takes the data from the photos in the database to make-up new details to fill into the larger file.

Let's Enhance: Low-res pixelated images into hi-res quality

Let’s Enhance is designed to be minimalist and extremely easy to use. It’s currently in its first version and will continually be improved based on user needs and feedback.

Start improving your pixelated images

After creating a free account, the website invites you to drag and drop your pixelated images. Once it receives your photo, the neural network goes to work, upscaling your photo by 4x, removing JPEG artifacts, and filling in the missing details and textures. You’ll need to wait a couple of minutes for the work to be done, but it’s worth the wait. You can try it for yourself here.

For another remarkable new image technology, check out Google’s watermark remover.


Source: https://petapixel.com/2017/10/31/website-uses-ai-enhance-photo-csi-style/