Sign Language Technology: Do's and Don'ts A Guide to Inclusive Collaboration Among Policymakers, Researchers, and End Users
Temas
Detalles
This white paper has arisen from the three-year SignON project , which received funding from the European Union's Horizon 2020 Research and Innovation Programme under Grant Agreement No. 101017255, and ran from January 2021 to December 2023. The project's primary objective is the development of an automatic translation application and service between spoken languages and sign languages. This application is the front face of a complex orchestration of tools, frameworks and models that jointly deliver the translation services. These rely on state-of-the-art machine and deep learning approaches and include sign language recognition, automatic speech recognition and natural language processing and understanding (NLP and NLU), to recognise and process users’ signed, spoken or written input, then translate it via a multilingual language model in order to, finally, produce signed or spoken (including written form) language output. The sign language output takes the form of a 3D virtual character, i.e. a virtual signer or a signing avatar, that performs the signs. However, SignON encompasses much more than just sign language technology. What sets SignON apart is that it is one of the first projects to actively engage in a co-creation process with the deaf and hard-of-hearing community. Following the lessons learned from this approach, this paper focuses on sign languages and the sign language users, and draws guidelines for an effective collaboration in which the deaf communities steer the research and development processes.