Generative Models for Scene Manipulation
Abstract
Automated localization and replacement of product images in wild scenes presents immense potential for delivering interactive experiences to people.
Advancements in deep learning research, GANs in particular, has extended research interest in the domain.