A team of researchers at the MIT-IBM Watson AI Lab developed an AI tool that is using a neural network to generate images almost identical to real photographs that are being uploaded by its user. It then lets the user remove any unwanted objects from those pictures or replace them with new ones. The tool is called GANPaint Studio and its demo version is available at ganpaint.io. The team acknowledges that their system could potentially be abused by bad actors, but they believe that more importantly, it will help researchers better understand how Generative Adversarial Networks (GANs) work to improve our ability at detecting fake images.
Watch Next
