A tech company, DataGrid, housed on the campus of Japan’s Kyoto University, has developed a deep learning algorithm which generates high-resolution, photorealistic images of people — faces, hair, outfits, and all — from scratch. Past attempts to create photorealistic portraits focused just on generating faces. These faces had flaws like asymmetrical ears or jewelry, bizarre teeth, and glitchy blotches of color that bled out from the background. This new system does away with all of the extraneous info that can confuse algorithms, instead posing the AI models in front of a nondescript white background and shining realistic-looking light down on them.
Watch the video in the link below showing off the technology, as the AI morphs and poses model after model as their outfits transform, bomber jackets turning into winter coats and dresses melting into graphic tees. The technology will soon be licensed out to clothing companies and advertising agencies interested in whipping up photogenic models without paying for lights or a catering budget. At the same time, similar algorithms could be misused to undermine public trust in digital media.
Each time scientists build a new algorithm that can generate realistic images that are indistinguishable from real photos, it also gives us a new warning that AI-generated media could be readily misused to create manipulative propaganda. Here’s hoping that this algorithm stays confined within the realm of fashion catalogs.