evidoria

← Back to browse

Good practice Imported

SignAll — AI Sign-Language Translation Co-Designed with the Deaf Community (Budapest, Hungary)

Hungary · Budapest · See the Hungary profile · See the Budapest profile

Evidence: Descriptive / self-reported Top 68% 56/100 · Ask Evidence Copilot about this practice

Budapest-founded SignAll built computer-vision AI that translates American Sign Language into English text in real time, developed together with Deaf linguists and Gallaudet University rather than for them. Piloted publicly at Gallaudet's Maguire Visitor Center from June 2018.

Details

Maturity
Pilot
Promoter
SignAll Technologies
Period
2016–ongoing (Gallaudet pilot from June 2018)
Region (NUTS)
HU110
Keywords
assistive technology, AI/computer vision, Deaf accessibility

Context

SignAll Technologies, spun off in Budapest, Hungary from the R&D centre of Dolphio Technologies, builds computer-vision AI that translates American Sign Language (ASL) into written English in real time, using multiple cameras and a depth sensor to track hand shape and movement.

Objectives

The company set out to build the system's linguistic core together with the Deaf community rather than for it, addressing the risk that automated sign-language translation tools are built without direct Deaf input.

Activities

SignAll partnered with Gallaudet University — the world's only university created specifically for Deaf and hard-of-hearing students — whose Deaf linguists and interpretation faculty helped build a sign database and validate the translation model; the resulting SignAll Chat product debuted publicly in June 2018 at Gallaudet's Maguire Visitor Center, letting Deaf and hard-of-hearing signers converse directly with non-signing visitors through a two-screen kiosk.

Results

Independent coverage (TechCrunch) and Gallaudet's own account confirm the pilot and the co-design process, though neither publishes translation-accuracy figures or data on how many other sites have since adopted the system.

Conclusions

The evidence base covers the design process and a single flagship deployment, not broad measured impact.

Implementation
Indicative cost
High (€500k–€5M) — The system requires multiple cameras and a depth sensor per installation plus ongoing AI model development, implying a hardware- and R&D-intensive cost profile relative to software-only accessibility tools.
Time to results
Long (> 3 years) — Development began around 2016, with the co-designed sign database and model built ahead of the June 2018 public pilot at Gallaudet, and the company's product line (SignAll Chat/Learn Lab) has continued since.
Staffing & skills
SignAll Technologies (Budapest-based, spun off from Dolphio Technologies' R&D centre), Gallaudet University Deaf linguists and interpretation faculty, including Dr Christian Vogler, director of the Technology Access Program

Conditions for success

  • Co-design embedding Deaf linguists and interpretation faculty in building the sign database and validating the translation model from early development
  • Partnership with Gallaudet University, the world's only university created specifically for Deaf and hard-of-hearing students
  • A public flagship pilot site (Maguire Visitor Center) to demonstrate real-world use

Common failure modes

  • No published translation-accuracy figures
  • No data found on adoption beyond the single Gallaudet Visitor Center pilot site
  • Confirmed deployment has not been shown to scale to additional locations

Where it fits

Governance type
private technology company partnered with an academic institution
Scale
single flagship pilot site
Income level
high-income

Commonly funded by

Own resources / municipal budget

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful