Neural Translation of Sign Language
Betreuer/in: Sarah Ebling, Mathias Müller
Introduction
The aim of this project is to conduct research into neural machine translation (NMT) between sign language and spoken language text. You will be involved in preparing an existing sign language dataset for use in NMT and performing NMT experiments.
Aim and Purpose
Important contributions will include:
- Deriving suitable representations for capturing the multilinear nature of sign languages
- Training NMT models using state-of-the-art frameworks
Requirements
Completion of the following lectures:
- Machine Translation
- Advanced Techniques of Machine Translation
Literature
Danielle Bragg, Oscar Koller, Mary Bellard, Larwan Berke, Patrick Boudrealt, Annelies Braffort, Naomi Caselli, Matt Huenerfauth, Hernisa Kacorri, Tessa Verhoef, Christian Vogler, and Meredith Ringel Morris. Sign language recognition, generation, and translation: An interdisciplinary perspective. In ASSETS 2019. ACM, October 2019.
Necati Cihan Camgöz, Simon Hadfield, Oscar Koller, Hermann Ney, and Richard Bowden. Neural Sign Language Translation. In Proceedings of CVPR 2018, 2018.