AI and Sign Language: Can Technology Bridge the Communication Gap?

Machine translation has come a long way for spoken languages, but sign languages have historically been left out of that progress. A recently completed European research project called SignON set out to change that, and the way it went about the work is just as notable as the technology itself.

What SignON Set Out to Do

SignON was a three-year project funded by the European Union’s Horizon 2020 program, running from January 2021 to December 2023, with a consortium of 17 European partners. The goal: build a free, open-source mobile application and framework capable of translating between multiple national sign languages and spoken languages, using video, audio, and text as inputs and outputs, including a customizable 3D avatar for sign output.

The project focused on five specific sign languages, Irish, British, Dutch, Flemish, and Spanish, alongside four spoken languages, but it was designed from the ground up to be extensible to new languages down the road.

Built With the Deaf Community, Not Just For It

What sets SignON apart from a lot of accessibility technology is how it was built. The project was intentionally co-created with Deaf community members from the start, rather than developed in a lab and handed off afterward. Davy Van Landuyt, project manager at the European Union of the Deaf (EUD), put it plainly: including Deaf people in a project isn’t enough on its own. They need to lead the research so the outcome actually reflects what they need.

That input directly shaped the final product, including more sophisticated avatar models designed to better capture the nuance of signed language, something earlier sign language avatars have historically struggled with.

What This Means for International Sign

SignON focused on national sign languages, but projects as it points toward a stronger possibility down the road: could similar co-created, Deaf-led technology eventually support translation involving International Sign as well? IS presents its own unique challenges for machine translation, since it has no fixed grammar or native signing community, but the co-creation model SignON pioneered offers a promising blueprint for how that kind of research should be approached, with Deaf leadership built in from day one.

The Bigger Takeaway

Technology alone doesn’t close communication gaps. Projects like SignON are a reminder that the people who actually rely on a language, in this case, Deaf and hard-of-hearing sign language users, need to be the ones steering how that technology gets built.

Interested in International Sign or the global Deaf community? Follow OneDeafWorld.com for more coverage of IS, accessible technology, and events from around the world.