Digital participation for deaf people.
Around 70 million deaf people live worldwide. Because written language is often a foreign language for them, only the dynamic presentation of content in sign language enables genuine digital participation. Digital accessibility is becoming increasingly important.
Using artificial intelligence and digital avatars, automated tools support the realisation of barrier-free communication.
Machine learning meets rule-based synthesis.
AVASAG develops a novel sign animation method for 3D avatars. By combining machine learning with rule-based synthesis methods, text is mapped onto signs — precisely resolving the temporal and spatial dependencies between sign elements.
Working with the sign language community, the project built a demonstrator for automated translation in the field of travel information, focusing on transport and tourism.
Project lead and core avatar expertise.
Charamel led the collaborative project as a specialist for interactive avatar-based assistance systems — together with yomma, Ergosign, DFKI, TH Köln – University of Applied Sciences and the University of Augsburg. Around 74% of the project volume was funded by the BMBF.