As the tech world witnesses a surge in AI tools and chatbots, an innovative AI imaging tool named DragGAN has stolen the limelight, offering a revolutionary breakthrough in the field of photo editing.
This model, first reported by The Verge, has made significant waves on social media, offering users the ability to mould and reshape images using just a mouse cursor and an intuitive graphical interface reminiscent of the widely-used Photoshop.
The powerhouse behind this technology is a team of researchers from Google, who in partnership with the Max Planck Institute of Informatics, are investigating a streamlined method to effectively control generative adversarial networks (GANs).
“The challenge,” according to the research team, “was to synthesise visual content that meets users’ needs, offering flexible and precise controllability of the pose, shape, expression, and layout of the generated objects.”
DragGAN: Photo Editing Made Easy
This is where DragGAN comes in.
Unlike traditional photo editors that primarily warp images, DragGAN takes a fresh approach.
It allows you to “drag” any points of an image to precisely reach target points, thus manipulating the pose, shape, expression, and layout of diverse categories such as animals, cars, humans, landscapes, etc.
The result is an image editing tool that allows even novices to move multiple points of an image along a desired path defined by the user, with the AI ensuring a realistic and natural final result.
“In essence, DragGAN deforms an image with precise control over where pixels go. The changes we make regenerate the entire object in the image, offering smoother edits and a more seamless editing experience,” the team explains.
A few examples showcased include transforming the shape of a car, changing the direction a lion faces, or even adjusting the height of a mountain.
Collaborative Interactions with Other AI Tools
What’s more, DragGAN can be used alongside text-to-image generative AI tools such as Midjourney or Runway.
When the results from these tools do not meet expectations, DragGAN can be used to edit them quickly and efficiently, going beyond the capabilities of even professional editing suites.
Future Prospects and Availability
While DragGAN is still currently in the research phase, the overwhelming interest in this tool has made waves on the internet, with the team’s homepage intermittently crashing due to high traffic.
However, the researchers have hinted that the code will be made available on their Github page in June 2023.
Although it’s not yet available for public use, the immense potential of DragGAN has already started questioning the way we think about image manipulation.
The surge of excitement around this tool indicates that we’re only just beginning to grasp the implications of this technology.
