For Edge browser users, Turing Image Super-Resolution will automatically convert online photos into high-resolution visuals.
Microsoft Bing introduced a new AI technology that would automatically enhance website photos and offer a 4K visual experience to webpages using Microsoft Edge. The Turing Image Super-Resolution technology allows images to be shown at a high resolution regardless of how terrible the source image is.
Microsoft’s Project Turing AI development team created the new technology.
Bing Maps Already Uses it
Bing Maps is already using the new technology to improve the quality of its satellite aerial photos.
Below is a comparison of aerial photography of Google’s Mountain View, California headquarters.
Microsoft’s Technology Development Process
Four key observations contributed to the model’s success.
Perceptual and GAN Loss Transformers for Vision: Enhance and Zoom Human Raters Noise Modeling
Human Raters
Microsoft recognized that the measures used to assess the effectiveness of image-related models did not correspond to human perception. As a result, they devised a side-by-side visual comparison tool that relied on human raters to assess the model’s effectiveness.
Noise Modeling
Microsoft used the method of beginning with high-quality photos, degrading them by introducing noise, and then teaching the model to restore the image to its former high-quality condition.
Perceptual and GAN Loss
This was done as part of the effort to make the outcomes more human-like.
According to Microsoft’s announcement:
“We discovered that refining our models merely based on pixel loss between output and ground truth pictures was insufficient to yield the best output that matched human perception”.
In response, we added perceptual and GAN losses as objective functions, and calibrated an optimal weighted mixture of the three losses.”
Transformers for Vision
Microsoft focused on the augment and zoom functions of Transformers, which were previously employed in language models.
This entails improving the image while also concentrating on scaling it up, which is a challenging task.
Shrinking a picture is usually straightforward. However, scaling up a tiny image usually magnifies the original image’s poor quality artifacts.
So the researchers devised a method for calculating and “recovering” missing picture data from a lower-resolution image and bringing it to a higher-resolution image.
Microsoft calls the process of scaling an image up, Deep Zoom.
Edge: 4K TV of Web Browsers
This new AI capability is envisioned by Microsoft as a method to bring a 4K visual experience to online browsing, as well as improving video meetings and family images posted to the web.
The technology is already available in Edge Canary, an experimental version of Edge.
Over the next several months, the new capability will be rolled out to the mainstream version of the Edge browser.
