New Image Editing Tool Uses AI To Dynamically Alter And Manipulate Images In 3D Using AI Regeneration

Artificial Intelligence (AI) has been rapidly altering the way content is created. A recently developed platform is offering simple controls and tools to manipulate and edit images.

A new platform, not yet available to the general public, is creating waves online. The promising AI model, called DragGAN is essentially an "Image Warping" tool, but its capabilities are mind-boggling. Let's look at the new tool and its features.

New AI Image Warping Tool Surfaces On The Internet

Several image generators that rely on AI to accept and understand text-based prompts have surfaced recently. A new tool allows created images to be manipulated. Although this may sound like Adobe Photoshop or other popular image editing tools, the infusion of AI makes a lot of difference.

The platform is essentially the result of a research paper, which has now transformed into a dedicated website. Accessing the powerful AI-based tool has been a challenging task primarily because the website is unable to handle the high traffic.

The website is intended to serve as a demo, and hence, it is impossible to fully evaluate or even appreciate the finer nuances and capabilities of the new tool. Nonetheless, AI-based image manipulation, and not just creation, shows the seemingly endless possibilities.

DragGAN Lets Users Drag Elements Of A Picture To Change Their Appearance

This model, named DragGAN appears like any other traditional image-warping tool. However, instead of simply smudging and mushing existing pixels, the model generates the subject anew with each edit.

As the researchers write: "Our approach can hallucinate occluded content, like the teeth inside a lion's mouth, and can deform following the object's rigidity, like the bending of a horse leg."

The team responsible notes what's really interesting about this work is not necessarily the image manipulation but the User Interface (UI). Content creators have been able to use AI tools like GANs to generate realistic images for a while now, but most methods lack flexibility and precision.

For quite some time now, users could tell an AI image generator to "make a picture of a lion stalking through the savannah," and they would obviously get one. However, the AI-generated image might not be the exact pose users want or need. This tool can help in tweaking or even altering the contents of the image to the intended parameters.