Introduction
Adobe, the renowned digital creative solutions company, is reshaping the landscape of image editing with its latest Photoshop Beta feature, dubbed “Generative Fill”. Integrated with Adobe’s Firefly AI, Generative Fill empowers users to modify photos with unprecedented flexibility and precision, setting a new benchmark in AI-based photo editing.
Table of Contents
Unleashing Firefly: The Power Behind Generative Fill
Adobe’s Firefly, a generative AI, now collaborates with the Photoshop core tools to provide a “deep integration” experience for users. Adobe’s answer to Midjourney, Firefly, was announced in March. Adobe has garnered praise for its responsible development of Firefly, opting to abstain from utilizing artists’ work to train the AI— a practice that has sparked controversy among its contemporaries.
Firefly forms the backbone of the new Generative Fill feature. It is a deep learning AI model, trained on millions of images from Adobe’s stock library. Its primary function is to associate imagery with text descriptions, enabling the feature to provide several AI-synthesized image options based on user text prompts.
How Does Photoshop Beta with Generative Fill Work?
Using Generative Fill is a simple, intuitive process. Users select an area of an existing image they wish to modify, following which a “Contextual Task Bar” appears. Here, users can input a description of the image they desire, like “a clown on a computer monitor”. Once this data is sent to Adobe’s servers, Firefly works to generate image options that align with the user’s description.
Generative Fill offers an AI technique called “inpainting,” a context-aware image generation method that seamlessly integrates synthesized imagery into an existing image. This provides multiple realistic options for the user, allowing them to iterate through a myriad of creative possibilities.
Moreover, Generative Fill allows for non-destructive image alterations, including additions, extensions, or removals. It automatically adjusts to the perspective, lighting, and style of the selected image.
Going Beyond Photo Manipulation
The Photoshop Beta with Generative Fill isn’t limited to photo manipulation. It also equips users with the power to erase objects from a scene or extend the dimensions of an image by generating new content surrounding the existing image. This technique, known as “outpainting,” has previously been featured in OpenAI’s DALL-E 2 image generator and editor.
Addressing Concerns: Content Authenticity and Regulation
Amid the excitement surrounding Photoshop Beta with Generative Fill, Adobe remains conscious of the potential misuse of such powerful technology. In response, Adobe blocks certain copyrighted, violent, and sexual keywords from generating content. Additionally, the company’s terms of use restrict people from creating “abusive, illegal, or confidential” content.
Furthermore, Adobe continues its commitment to the Content Authenticity Initiative, applying Content Credentials to add metadata to generated files, aiding in tracking their provenance.
Availability of Generative Fill
Currently, Generative Fill is available to all Creative Cloud members with a Photoshop subscription or trial via the “Beta apps” section of the Creative Cloud app. While not yet accessible for commercial use, not available to individuals under 18, and not available in China, Adobe plans to make Generative Fill accessible to all Photoshop customers by the end of the year.
Those without a Creative Cloud subscription can also test Generative Fill for free on the Adobe Firefly website with an Adobe login through a web-based tool.
Conclusion
Adobe’s launch of the Generative Fill feature in Photoshop Beta marks a significant stride in the realm of AI-powered photo editing. The blending of Adobe’s Firefly AI and Photoshop’s core tools creates a powerful and intuitive user experience. As AI continues to redefine creative boundaries, Adobe’s Photoshop Beta with Generative Fill offers a glimpse of the future of digital image manipulation.
