Adobe Is Not Done With Its Generative AI Offerings

Adobe is ramping up its commitment to generative AI, releasing a slew of new tools within its Firefly suite, whose products have generated more than 3 billion images since March, according to the company. During its annual creativity conference, MAX, held in Los Angeles, the creative software provider introduced Firefly Image 2 and Firefly Vector, among other features. These additions to its creative generative AI platform were trained on licensed content such as Adobe Stock. Brands like Mattel, NASCAR and IBM have adopted the Firefly suite into their creative processes to expedite workflows. LinkedIn Introduces Generative AI-Powered Tool for B2B Marketers "Adobe’s creative tools continue to play an important part in our design processes at Mattel, including on beloved brands like Barbie and Hot Wheels,” Chris Down, evp and chief design officer at Mattel, said in a statement. “Adobe Firefly has enhanced our workflows with trusted capabilities, driving inspiration and productivity from ideation to execution." Firefly's Image 2 introduces creative controls that enhance image output quality, while Firefly Vector speeds up tasks such as finding inspiration, creating mood boards, and generating marketing and graphics, including product packaging, brand logo designs and website development. “We see a content supply chain problem for marketers who want to create more content in a more hyper-personalized way,” said Alexandru Costin, vp of generative AI at Adobe. “[Adobe’s generative AI] capabilities [enable that] by making creative professionals more productive and creative while they explore more variation, faster." Finer details, higher image quality The advanced creative controls in Adobe's Firefly Image 2 offer improved image quality and precise rendering of human subjects, such as capturing intricate details like hair and pores. The NextGen image modeling tool helps people refine their prompts to generate images that better align with their creative vision. A feature called Generate Match lets people upload their own reference images to Firefly Image 2, letting it mimic specific image styles and incorporate them into its output. This feature helps artists and marketers adhere to their brand guidelines. “[Adobe] doesn’t take the content of the image but only the style,” said Costin. These new developments were prompted by feedback from consumers using the first generation of Adobe Firefly. Comments ranged from requests for improvements such as higher resolution capabilities and the option to upload images for style emulation, both of which aimed to enhance users' productivity and creative experience. This was gathered across Adobe's Discord community of half a million members, as well as through the rating system on the Firefly web application. The new tools were crafted with feedback from Adobe's community. Inspiration and mood board creation Marketers’ creative processes such as mood boards and product mock-ups are getting a generative AI lift with Firefly Vector. Adobe is using the gen AI capabilities of Adobe Illustrator through Firefly Vector, currently in beta. The Text to Vector Graphic feature within Illustrator generates a wide array of editable vector graphics through text prompts, a feature similar to Adobe's Generative Fill in Photoshop. Adobe’s Hilarious ‘Office Space’ Sendup Starring Hasan Minhaj Is a Tribute to 30 Years of the PDF The new model was trained on licensed content, such as Adobe Stock, and public domain content where the copyright has expired, and generates high-quality vector output specializing in scenes, icons and patterns. Built-in tools let people capture images of physical objects, such as bottles or T-shirts, and apply vector art, such as logos, onto these objects with precision and ease. “It could be a T-shirt with creases and [the tool] will understand that and do a good job of placing a logo on a 3D object,” said Costin. Tools within Firefly Vector can mock up products.