Visipedia founders awarded 2025 Stibitz-Wilson Award for pioneering work in human-centred AI

Established in 1997, the Stibitz-Wilson Awards honour innovators whose contributions have advanced the fields of computational and biological sciences. The award was presented on 26 September 2025 at The Ellen Theatre in Bozeman, Montana, during the annual Stibitz-Wilson Awards Ceremony hosted by the American Computer & Robotics Museum (ACRM). This year’s theme is “Code for Tomorrow – Innovating for Classrooms, Communities, and Conservation.” 

A collaboration rooted in mentorship

The seeds of Visipedia were planted in the early 1990s, when Belongie began conducting research as an undergraduate student under Perona’s mentorship. What began as an academic relationship evolved into a long-term collaboration that laid the foundation for Visipedia - a project that continues to shape the future of responsible AI innovation.

“Visipedia was born out of a shared belief that AI should amplify human expertise, not replace it,” says Serge Belongie. “We’re honoured to see this vision recognised, and grateful to the communities who have helped bring it to life.”

AI that grows with communities

Visipedia is an academic project that explores new machine learning techniques and systems designed to empower communities of experts. It leverages crowdsourced expertise and deploys it in practical, scalable ways, creating a self-reinforcing system of knowledge that grows stronger with community engagement.

At a time when mainstream AI approaches often sidelined human input, Visipedia demonstrated how technology can be designed to collaborate with people, enabling them to contribute meaningfully to complex visual classification tasks.

Visipedia’s impact is evident in its contributions to large-scale, community-driven datasets such as eBird and iNaturalist. Through these collaborations, Visipedia has helped build and continuously refine global image datasets that support automated species identification – most notably through tools like Merlin Bird ID and the Seek app. These systems not only assist users in recognising biodiversity but also strengthen the capacity of expert communities by feeding back improved models and insights.

This approach, where machine learning is used to harness and amplify crowdsourced expertise, continues to serve as a conceptual foundation for the moonshot projects at the Pioneer Centre for AI, which aim to push the boundaries of inclusive, human-centred artificial intelligence.