ML Super Resolution has been trained to intelligently preserve details by identifying specific features in an image, like colors, textures, edges, patterns, gradients, and upscaling each individually.
Remove Noise and Artifacts
Small images often feature noise and compression artifacts, which ML Super Resolution removes during the upscaling process — not only enlarging the image, but also improving its quality.
Compared to traditional algorithms, ML Super Resolution requires between 8 to 62 thousand times more processing power. Thanks to Core ML and the latest Mac graphics processors, images are upscaled in just a few seconds or less.
Use ML Super Resolution to upscale RAW images while fully preserving their extended range data.