This system examines pictures of human faces to identify and digitally erase makeup. It employs an autoencoder, a type of AI, to learn what makes faces attractive and then assesses facial beauty. The system is trained using images of individuals showing their face both with and without makeup, and it can classify makeup types and pinpoint where it's applied on the face.
Why it matters: Filed before deep learning, especially generative models like GANs, revolutionized realistic image manipulation and beauty assessment. The accuracy and accessibility of such AI systems for digital makeup removal and attractiveness analysis have vastly improved since 2015.
AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.
Reinvent this with AI