Artificial intelligence (AI) is generating higher quality images on the web.
Take Mountain View, Calif.-based Google’s RAISR Rapid and Accurate Image Super-Resolution technology that was introduced last November.
It uses machine learning to produce high-quality versions of low-resolution images. Google is now integrating RAISR with its online services to upscale large images on Google+ and save users data bandwidth in the process.
When a user requests an image, Google+ retrieves a version that’s a quarter of the size, and uses RAISR’s algorithms to “restore detail on [the] device."
Google+ Product Manager John Nack explained in a post about RAISR the technology also improves site load time and up to 75 percent less bandwidth per image.
Google claims that RAISR produces results that are comparable or better than currently available super-resolution methods, and does so roughly 10 to 100 times faster.
It works in a similar way to most up sampling methods — inserting new pixels into low-resolution images to make up for lost detail. But while traditional sampling uses fixed rules to work out which new pixels to use where, RAISR adapts its methods to each image.
Google is applying RAISR to more than 1 billion images per week and plans to roll this technology out more broadly.
Twitter is also using AI with images. Last year, Twitter bought UK AI startup Magic Pony, a London-based firm that developed machine learning techniques for visual processing.
Calling machine learning "the core of everything we build at Twitter,” CEO Jack Dorsey blogged that the buy builds on other investments, including acquisition of Madbits in July 2014 and Whetlab in June 2015.