So far, we've seen simple images animated with artificial intelligence (AI). In the next step, Google has successfully experimented with making high-quality, low-resolution (low-resolution) images clear and high-quality. Some of the images even have missing eyes and other impressions, but the Google algorithm has done it beautifully.
Google has provided details and some pictorial examples of this technology in its blog. Google has a separate AI division that is active in many fields. This new technology has been dubbed the 'High Fidelity Image Generation Using Diffusion Model'. Thanks to this, old and grainy pictures of the family can be made completely new.
The same technology can be used to visualize CCTV videos and images. On the other hand, raising the standard of medical imaging will go a long way in identifying the disease and saving the lives of patients. The diffusion model makes the image clearer, clearer and clearer in just a few seconds.
According to Google, the technology uses artificial intelligence to reduce noise (ie grains) from images, while AI explains how the components of the image will be formed and thus a clear image emerges. Google has dubbed it Super Resolution by Repeated Refinement, which stands for SR3.
Thanks to this, it is surprising to see altered images because the altered images become much clearer than the original, as if a blurred image clarifies the original and natural image. Sometimes difficult-to-recognize faces come to the fore.
