Using sign language technology to make public health information more accessible
This project aims to enhance the accessibility of health information for deaf citizens by translating online information about common health issues and medications into sign language using machine translation and avatar technology. Health information in written form is often poorly accessible for deaf individuals whose primary language is a sign language (to understand this as a hearing person, consider how challenging it would be to learn to read in a foreign language with an unfamiliar alphabet, e.g. Thai, without ever hearing how the letters in the alphabet, มีขคดีอี etc., are pronounced). To address this issue, the project will establish a new dataset containing health content in multiple formats (Dutch text, videos featuring human signers, and 3D animations featuring signing avatars) and then use machine learning techniques to create an automated translation system. The signing avatars, powered by motion capture and 3D animation technology, will be designed to simulate human native signers, ensuring that the translations are not only accurate but also culturally and contextually appropriate.
Key areas of technological development include improving the fluidity and realism of the avatar’s movements to make it possible to convey the nuances of health information in sign language. The project will place emphasis on ensuring that the translations are medically accurate, compliant with regulatory standards for health information, and accessible for individuals with different levels of sign language proficiency.
User testing and feedback are integral to the development process. Input from target users will ensure that the system meets their needs and preferences. By providing sign language translations of critical health information, the project seeks to empower deaf individuals to make informed decisions about their health, promoting health equity and inclusivity in healthcare services, while also contributing to reducing the collective costs of the healthcare system.
