Papers by Richard Bowden
Select and Reorder: A Novel Approach for Neural Sign Language Production (2024.lrec-main)
Copied to clipboard
| Challenge: | Sign languages face significant challenges in achieving accurate translation due to the scarcity of parallel annotated datasets. |
| Approach: | They propose a method that breaks down the translation process into two distinct steps: Gloss Selection (GS) and GlosSelection (GR) they use non-autoregressive decoding to achieve faster inference speeds and reduced computation . |
| Outcome: | The proposed method achieves state-of-the-art BLEU and Rouge scores on the Meine DGS Annotated dataset, demonstrating a substantial improvement of 37.88% in Text to Gloss (T2G) Translation. |
SMILE Swiss German Sign Language Dataset (L18-1)
Copied to clipboard
Sarah Ebling, Necati Cihan Camgöz, Penny Boyes Braem, Katja Tissi, Sandra Sidler-Miserez, Stephanie Stoll, Simon Hadfield, Tobias Haug, Richard Bowden, Sandrine Tornay, Marzieh Razavi, Mathew Magimai-Doss
| Challenge: | The goal of an ongoing three-year project in Switzerland is to pioneer an assessment system for lexical signs of Swiss German Sign Language (Deutschschweizerische Gebärdensprache, DSGS) that relies on sign language recognition. |
| Approach: | The goal of the project is to pioneer an assessment system for lexical signs of Swiss German Sign Language that relies on sign language recognition. |
| Outcome: | The system will give adult L2 learners of DSGS feedback on the correctness of the manual parameters (handshape, hand position, location, and movement) of isolated signs they produce. |