AI bracelet offers real-time sign language translation for text and voice

Researchers at the Insight SFI Research Centre for Data Analytics based in DCU have developed a wearable sensor that can read and translate Irish Sign Language (ISL) in real-time.

The device, worn on the arm, reads electrical signals in the muscles that are activated when the wearer is using ISL, in a technique known as electromyography.

The technology will be showcased at the International Conference on Content-Based Multimedia Indexing (CBMI), taking place in Dublin City University from September 4 to 6 2019. The event brings together researchers in content-based multimedia indexing for retrieval, browsing, management, visualisation and analytics.

This year’s event is hosted by the Insight SFI Research Centre for Data Analytics, which has a network of projects underway in the area of wearable technology to monitor health and augment human performance.

This latest project has been developed by researcher Laura Galea under the supervision of Professor Alan Smeaton at Insight DCU.

“We anticipate that a workable, wearable device will be available to users within about two years,” says Professor Smeaton. “Right now the device can translate ISL letters into text or voice with up to 90 per cent accuracy. We are working on an autocomplete function, as you would use in texting, to increase that accuracy rate.”

The Insight SFI Research Centre for Data Analytics is a 450 strong research network based across four lead universities – DCU, NUI Galway, UCC, UCD – developing technologies in the areas of the Augmented Human, Smart Enterprise and Sustainable Societies.