Quickly, the mannequin showcasing your on-line clothes purchases might not have truly made the pose within the image. That’s as a result of a neural community can now repose human beings and alter their garments in pictures with out shedding key particulars.
Badour AlBahar at Virginia Tech in Blacksburg and her colleagues at Adobe Analysis developed an algorithm that breaks down a supply picture into constituent physique components, with a neural community figuring out the place key joints and limbs are. It’s then fed the goal pose of how the person desires the mannequin to face, earlier than it identifies the brand new positions of …
