Interactive dresses to achieve "instant dressing" Adobe + AI collides with new sparks丨Technology observation

Cover News trainee reporter Ma Xiaoyu

With interactive skirts, motion quality fixes, fast 3D modeling tools, and seamless scene compositing, Adobe gives you a "sneak peek" of future image generation.

"Project Primrose" makes it easy to change the look and design of clothing

Recently, at the 2023 Adobe Max conference, Adobe research scientists and engineers demonstrated for the first time several cutting-edge experimental technologies that may one day become features of Adobe products. Much of the content of this presentation can also be experienced directly in the official or beta version, rather than a mere demo conference.

As a well-known software company, Adobe focuses on multimedia creation and creative software products, common Photoshop (PS), Premiere (PR), Audition (AU) and other design software, are Adobe products. It leverages generative AI to give creators innovative tools across multiple media, quickly bringing these designs into real-world objects at the click of a button, taking creativity to a whole new level.

An interactive dress made of textiles that changes patterns in an instant

Among this year's many AI innovation projects, the most notable is "Project Primrose", an interactive dress that can change design and style in an instant. Made from a wearable and flexible non-luminescent textile, users can use software such as Adobe Firefly, Adobe After Effects, Adobe Stock, and Adobe Illustror to create content and project it onto the entire surface of the scaled dress, quickly adding animation and effects to their designs.

“Project Primrose ”

Adobe research scientist Christine Dierk led and demonstrated the project, with the stage wearing an interactive dress covered in scales that changes patterns at the push of a button. Not only that, but the dress is also able to sense the wearer's movements, allowing the pattern to sway and flow with the wearer's movements.

Project Primrose is a non-luminescent material that is highly responsive to external factors and can be used as a canvas for new designs or low-power text ads. In addition, the material can be cut into any shape and is able to diffuse light dynamically. Researchers have already applied the technology to fabrics like handbags, and with the release of interactive skirts.

Dierk revealed that before the Adobe MAX 2023 demo of the interactive skirt, her research team published research on reflective light diffusion modules for non-luminous flexible display systems in 2022.

Full of "black technology", Adobe + AI collided with new sparks

In addition to interactive clothing, Adobe Max showcased dozens of AI innovations such as motion quality restoration, automatic image generation, custom fonts, reflection cancellation, fast 3D modeling tools, seamless scene compositing and more, transforming labor- and cost-intensive processes into processes that can be quickly generated in minutes. All the user has to do is upload the relevant file, press the generate button.

"Project Poseable" turns a 3D model into a 2D AI image

Take the automatic translation project Project Dub, which uses generative AI to automatically voice video or audio clips in more than 70 languages and more than 140 dialects. It uses speech-to-speech to automatically translate and match the speaker's voice, tone, rhythm, and acoustics of the original video, whether the video or audio clip is brand new or from a video clip from the user's video archive.

Another example is the automatic generation of the project "Project Fast Fill", the power of this tool is that it allows AI repairs and AI generation to be applied to movies, and the effect is almost perfect. It can remove the background character of the film with one click, add a man's suit and tie, and change the latte shape of the coffee in the picture.

Project Fast Fill applies AI generation to movies

You can even imagine that with just a few lines, AI can help you automatically draw patterns. The "Project Draw And Delight" project can help you achieve this effect, through the simple sketch drawn by the user, let the AI recognize and generate more refined and complete patterns, and even automatically color, which is simply the gospel of the "handicapped party".

"Project Draw And Delight" turns simple sketches into more elaborate and complete patterns

Go all in AI! Get more people involved in creative production

Adobe announced three new generative AI models on October 11, adding powerful capabilities to Illustrator and Adobe Express, and dramatically improving Photoshop's text-generated image capabilities.

At the Adobe Max event, Adobe unveiled the Firefly Image 2 model, the latest version of the original Firefly AI image generator that supports popular features like Photoshop's generative fill, as well as two new Firefly models for generating vector images and design templates.

Adobe says the Firefly Image 2 model produces significantly higher quality images than its predecessor, especially when rendering realistic human bodies, involving high-frequency details such as leaves, skin texture, hair, hands, and facial features. Images produced with the Firefly Image 2 model have higher resolution and have more vivid colors and color contrast.

"Project Fast Fill" removes the background characters of the film with one click

The Image 2 model also introduces new AI editing capabilities to help users customize the results. Photo settings can manually or automatically adjust the depth of field, motion blur, and field of view of the resulting image, just like manual camera controls. A "Prompt Guide" feature has also been added to help users improve the wording of text descriptions and autocomplete prompts for efficiency.

Firefly Image 2 is available today through the Firefly beta and will soon be available in the Creative Cloud app. It is not currently available in Photoshop (Standard, Beta, or Web).

The combination of Adobe and AI has generated a lot of new sparks, and Adobe also said that it will continue to introduce more features in the coming period. It believes that the update and development of tools is not to raise the threshold of creation, but to allow more people to participate in it.

Although Adobe has proved its strength with more than a dozen projects in terms of innovative technology, it is inevitable to fall into the "strange circle" of AI development and technology abuse.

Scott Belsky, head of design at Adobe, said the company has developed "new policies and safeguards" to protect generated matches from abuse, and that the feature will prompt users to agree to Adobe's terms of use and confirm that they have permission to use uploaded images. Generating matches will also remain tested while the Company seeks feedback, during which time users will not be allowed to use them for commercial purposes.

Still, there's currently little to really stop users from imitating protected content, which could create a bigger divide between Adobe and creatives opposed to AI copying its style. At the moment, the system seems to be more about limiting Adobe's liability than preventing imitation in the first place.