Abu Dhabi patent backs sign language AI

Abu Dhabi University said an associate professor and a former student have been granted a patent for an artificial-intelligence application designed to translate between sign language and spoken or written language in real time, marking a push by the institution into assistive technology with a clear accessibility focus. The university identified the inventors as Dr Modafar Ati, associate professor of computer science and information technology, and alumna Reem Al Bostami, who graduated from its BSc in Information Technology programme. The patented platform, called HearMe, is intended to support people with hearing impairments in classrooms, training settings and workplaces.

According to the Abu Dhabi Media Office, HearMe is designed to work in two directions: converting signed gestures into written words and turning typed text into animated sign language. The same announcement said the platform also supports multilingual translation across different sign systems, naming American and French sign languages among those covered. That gives the project a broader ambition than many single-language assistive tools and places it within a growing race to build systems that can serve users across borders rather than in one linguistic setting alone.

Ati said the goal was to ensure advances in AI translate into “practical solutions” with measurable social impact, while framing the patent as both a technical milestone and an extension of the university’s inclusion agenda. The Abu Dhabi Media Office linked the application to the UAE’s National Policy for Empowering People of Determination and said the technology is meant to widen participation in higher education and professional training. Abu Dhabi University also tied the development to its wider Vision 2027 strategy and to earlier inclusion initiatives carried out with partners including the Authority of Social Contribution – Ma’an and the Zayed Authority for People of Determination.

The announcement lands as assistive communication tools attract greater attention from universities, start-ups and public bodies trying to use AI for practical social applications rather than only consumer automation. The World Health Organization says more than 1.5 billion people worldwide live with hearing loss and 430 million have disabling hearing loss, underlining the scale of the communication gap such tools aim to address. WHO also says the global economic cost of unaddressed hearing loss is close to $1 trillion a year, giving policy makers and developers a strong incentive to back technologies that improve access to education, employment and public services.

Yet the patent announcement also arrives in a field where technical claims are often easier to make than to prove at scale. Academic reviews and dataset papers published over the past two years have highlighted persistent weaknesses in sign-language AI, including limited datasets, uneven language coverage and difficulties handling continuous signing, context and regional variation. Researchers have also warned that many systems are trained disproportionately on a small number of well-resourced sign languages, raising the risk that tools marketed as inclusive may perform unevenly in real-world multilingual settings.

That makes the next phase of HearMe at least as important as the patent itself. The Abu Dhabi statement presents the technology as a working platform for academic and professional environments, but it does not spell out benchmark accuracy rates, deployment timelines, external validation results or whether interpreters and Deaf community representatives were involved in testing. Those details will matter if the application is to move from a protected invention to a trusted tool used in live settings where mistranslation can carry educational or professional consequences. European Deaf advocacy guidance on AI and sign language has also argued that such systems should remain explainable and open to human review, especially in higher-risk uses.