This project aims to transcribe in SignWriting a large vocabulary of single-sign videos in various signed languages.
Allow the matching of existing entries in SignBank to videos in SpreadTheSign. This should be less time-consuming than transcription from scratch, and should allow a baseline transcription system to be developed.
For every video (in the database, or uploaded), the system should predict a transcription, which could be followed by human post-editing.