International Workshop on Spoken Language Translation (IWSLT) 2010

Paris, France
December 2-3, 2010

Sign Language Machine Translation Overkill

Daniel Stein, Christoph Schmidt, Hermann Ney

Human Language Technology and Pattern Recognition, RWTH Aachen University, Germany

Sign languages represent an interesting niche for statistical machine translation that is typically hampered by the scarceness of suitable data, and most papers in this area apply only a few, well-known techniques and do not adapt them to small-sized corpora. In this paper, we will propose new methods for common approaches like scaling factor optimization and alignment merging strategies which helped improve our baseline. We also conduct experiments with different decoders and employ state-of-the-art techniques like soft syntactic labels as well as trigger-based and discriminative word lexica and system combination. All methods are evaluated on one of the largest sign language corpora available.

Full Paper

Bibliographic reference.  Stein, Daniel / Schmidt, Christoph / Ney, Hermann (2010): "Sign language machine translation overkill", In IWSLT-2010, 337-344.