Home App Adobe and Nvidia both launch creative generative image AI to to tackle...

Adobe and Nvidia both launch creative generative image AI to to tackle copyright issues

0

Adobe and Nvidia released new tools that use artificial intelligence to generate images, while also solving two of the most difficult legal issues when using the technology, copyright and payment.

Adobe has added artificial intelligence technology to some of its most popular software, including Adobe Photoshop and Adobe Illustrator, to speed up the process of image generation and text effects. Creators of works used by these tools will be able to get paid, the company noted.

Adobe has introduced Adobe Firefly, its collection of creative generative artificial intelligence models, to users. First release focused on image generation and text effects, will be integrated directly into Adobe Creative Cloud, Document Cloud, Experience Cloud and Express Cloud workflows, the first apps to integrate with Firefly will be Adobe Experience Manager, Adobe Photoshop and Adobe Illustrator.

Nvidia has launched its own service called Picasso, which uses artificial intelligence to generate images, videos and 3D applications from text descriptions. Nvidia has technically trained and plans to pay royalties on images licensed by Getty images, Shutterstock Inc, and Adobe.

The move is a milestone in an ongoing tension between copyright holders’ rights and emerging technologies. Image generation techniques are “trained” on billions of images, but it’s unclear whether it’s legal to use them this way.

Earlier this year, Getty Images sued UK-based open-source artificial intelligence company Stability AI, creator of the open-source art generator Stable Diffusion, for copying more than 12 million images from its database without permission.

“This collaboration (with Nvidia) demonstrates the viability of responsible AI development paths and the uniqueness of Getty Images content and data,” said Getty Images CEO Craig Peters. Peters told the media in an email.

“This aligns with our belief that generative AI is an exciting tool that should be built on permissioned data, visuals, and individual privacy.”

A new Adobe AI-enhanced feature called “Firefly” allows users to describe the images, illustrations or videos their software will create with words. Because the artificial intelligence has been trained on Adobe stock images, publicly licensed content and stale content whose copyrights have expired, the resulting creations are safe for commercial use, the company said.

The company is also advocating for a generic “no training” label, so photographers can request that their content not be used to train models.

“We’re very interested in making this creator friendly,” Ely Greenfield, Adobe’s chief technology officer for digital media, told the press.

If an Adobe user asks the system for an image in the style of a particular artist, “it’s not going to generate an image that mimics that person’s style,” Greenfield said. “As an artist, you can promote it. If someone wants to use your style, You can actually sell the client the right to use your style.”

Nvidia’s Picasso AI Image Generator is part of a series of AI cloud products announced at the GTC developer conference.

“That’s where the fundamentals of what the market is interested in coming out,” Greg Estes, Nvidia’s vice president of developer programs, said of working with partners like Getty.

“Because of other software providers or any kind of business, they don’t want to participate in (image-generating AI technology) because they don’t know the source of the underlying training images,” he said.

Zhu Junyan, an assistant professor at Carnegie Mellon University’s Robotics Institute, said it’s not uncommon to train open-source AI models on billions of images. He added that many factors determine whether a photographer knows their work is being sampled, including whether the photographer is famous or whether the training dataset is publicly available.

Zhu Junyan said that it is hoped that photographers and artists can eventually benefit from using this technology to empower their artistic style.

“Content creators’ livelihoods depend on respect for intellectual property and the value of their creative endeavors,” Getty’s Peters said.

“We believe that innovation and creativity can thrive in an environment where artists, photographers, videographers and creatives are paid fairly, especially when their work is used for commercial purposes.”

Exit mobile version