New AI technology reconstructs damaged images, video in real-time

Global research and development firm Cambridge Consultants has developed an artificial intelligence (AI) software able to reconstruct images in real-time that have been previously damaged or obscured, according to a Dec. 4 report by Venture Beat.

The AI system—called DeepRay—utilizes Adobe’s smartphone distortion correcting system for selfies and a technique from Nvidia that can fix images or live video that are corrupt and contain holes, according to the article. The technology may provide healthcare providers with more accurate medical imaging data, according to a prepared statement from the company.

Specifically, DeepRay utilizes a generative adversarial network (GAN)—a machine learning based architecture—to invent video scenes while attempting to remove any distortion from a frame. The technology will be showcased at the 2019 Consumer Electronics Show in Las Vegas, according to the article.  

“Never before has a new technology enabled machines to interpret real-world scenes the way humans can—and DeepRay can potentially outperform the human eye,” Tim Ensor, commercial director for AI at Cambridge Consultants, told VentureBeat. “The ability to construct a clear view of the world from…video, in the presence of continually changing distortion such as rain, mist, or smoke, is transformational.” 

See Venture Beat’s entire article below.