Submitted by mitedu6137 in technology
A new framework known as MAGE merges the two key tasks of image generation and recognition, typically trained separately, into a single cohesive system. The work was developed by researchers from MIT and Google.