Adobe Firefly

Community Gallery

Overview

Firefly Community Gallery

The Firefly Community Gallery offers a platform for Firefly users worldwide to submit and share their AI-generated art, images, and videos in a real-time feed with other users.

The Gallery is designed to inspire creativity by showcasing what other creatives are making with the generative AI tool and inspire other users to use these works as a starting point for their own creations.

ROLE

Lead Designer

  • Define MVP scope & strategy
  • Research
  • User Journey and Wireframes
  • Visual Design
  • User Testing
  • Development specs
Team

Phil/Anum - Community Design & Strategy
David Wang - Project Manager
Marisa - Content Strategy
Adobe Firefly team
(PM's, Designers, Research, Engineers)

STRATEGY

What are we trying to accomplish?

From a business standpoint:
To further showcase the Gen AI capabilities of Adobe Firefly, create a community platform where users can discover inspiration from a range of AI-generated images, text effects, and other visual content made by fellow community members using Adobe Firefly. This platform should inspire and encourage users to share their own creations using the tool's features. The gallery should demonstrate the potential of Adobe Firefly in enhancing creativity and showcase the various artistic possibilities achievable with the application.

From an experience standpoint:
Empower Firefly users with more control to learn, discover, be inspired, and remix the work of others, igniting new ideas they may have never imagined.

RESEARCH

Methodology

Virtual user interviews

The research team conducted online virtual interviews with various users from different creative backgrounds, both current and new users, as well as users of other platforms similar to Adobe Firefly.

  • Current users:
    the team surveyed current and active users of Adobe Firefly.
  • New users:
    the team also surveyed new users of Adobe Firefly.
  • Users of similar platforms and tools:
    as well as interviewing users of competing platforms.

Competitive research

After conducting the virtual interviews, we identified popular comparable market platforms that are widely used for various creative purposes. We analyzed and examined the functionalities of these platforms and identified areas that align with our current focus and efforts.

Interview Insights

Current gallery feedback

Some of our initial assumptions were validated. Users were uncertain whether the gallery images were created in Adobe Firefly or by other users. Users have also expressed the desire to be able to submit their own creations and ability to filter the gallery based on preferences.

Reasons for using Adobe Firefly or similar Gen AI platforms

  • Browsing simply for fun
  • Curiosities of Gen AI
  • Searching for image styles as a starting point
  • Collecting inspiration for specific types of projects

Competitive Insights

We assessed four different platforms similar to Adobe Firefly, as well as Pinterest, which shares a similar framework with our subjects. We then identified specific areas within the platform experience that directly address our objectives and ranked them, with 1 being the strongest, 2 in the middle and 3 as the least strong.

After analyzing different areas of experience, we found that each of the four direct competitors had specific strengths, but none were common across all of them. Leonardo came the closest to meeting our objectives regarding experience, while Pinterest, an outlier of the group with framework similarities, ranked the best in all categories.

We’ve determined that certain experiences were intentionally suppressed to motivate users to explore and unleash their imagination through the capabilities of the next-generation AI tool instead of simply finding what others have already created.

APPROACH

Design Principles

In our research, we discovered that one of the challenges that users were facing was the difficulty in narrowing down their search for inspiration from the gallery. While some users initially found the infinite scrolling fun for browsing, many others felt that it was a waste of time and hindered their intended use of the gallery.

Overall, the main challenge is to assist users in finding the right image to start their exploration.

To effectively tackle these challenges, we have identified the following capabilities for the gallery.

  • A search functionality designed to locate a particular type of generated image.
  • A filtering feature that enables users to narrow their search to specific categories.
  • Over time, curate a personalized feed to help users discover content tailored to their interests based on interactions.
  • Provide a seamless browsing experience that enables users to access an image and easily return to the feed quickly.
  • Furthermore, the ability to share recommendations similar to images with matching attributes.
DESIGN

Search & Filtering

One area of focus we targeted to address these problems was to implement a search functionality that combs through text prompts that describe the images submitted to the gallery.

The search field provides prompt suggestions as the user types their query.

GALLERY SEARCH FIELD

As part of our early explorations, we included a comprehensive filter functionality similar to the options in the AI editor. However, I felt that it was too heavy of a feature for such a new product that we have yet to learn from and validate its use.

FULL-ON FILTERING

I’ve designed a more streamlined filter that is adequate for its intended use and ideal for early adopters to learn from and iterate upon.

The idea is to show tags related to the image when hovering over it, providing useful information for users browsing the gallery. Clicking on each tag also allows users to filter the gallery to display images with those specific style attributes.

Another component that allows for personalization in the feed is the “like” feature included in the image hover. The gallery collects these “liked” images from the users and creates a curated collection of similar images to personalize the users’ feeds based on these likes.

MVP FILTER COMPONENT

MVP FILTERING

Ease of Exploration

Early in our user interview research, it became apparent that users visit the gallery to browse, learn, and draw inspiration, sparking their imaginations. This journey requires a smooth and efficient process, making it simple for users to browse the feed, view images, learn more, and repeat.

I’ve designed an experience that allows users to view images in full detail through an overlay. Users can also take full control of the editor, remix the image to create their own version, and then seamlessly return to the feed, repeating the process.

OVERLAY EDITOR

Lastly, while viewing an image in detail, users can also browse a curated feed of similar recommended images, helping them narrow down their search for inspiration.

Product Unveiling

In late 2023, Adobe unveiled the new Firefly Community Gallery for users worldwide. The platform has inspired and enabled users to easily submit and share their AI-powered artwork, images, videos, and other projects, resulting in over 5 billion images generated and counting.

The launch of the new Firefly Community Gallery included just enough features to attract early adopters and gather insights for our next steps.

REFLECTION

Experience and Growth

The project came together on short notice and progressed pretty quickly. Integrating the Community Gallery within the Adobe Firefly product required extensive collaboration from the Community and Firefly team, researchers, engineers, and other contributors from different sectors to ensure a seamless experience.

Despite encountering numerous challenges, we managed to develop most of the features we released with our limited resources. As a designer, I had to adapt quickly to create solutions that fit our circumstances and make well-informed decisions on the fly. Designing a new product in a fast-paced environment with very limited resources was a valuable learning experience.