Sign Language Transformers: Joint End-to-End Sign Language Recognition and Translation

Citations Over TimeTop 10% of 2020 papers

Abstract

Prior work on Sign Language Translation has shown that having a mid-level sign gloss representation (effectively recognizing the individual signs) improves the translation performance drastically. In fact, the current state-of-the-art in translation requires gloss level tokenization in order to work. We introduce a novel transformer based architecture that jointly learns Continuous Sign Language Recognition and Translation while being trainable in an end-to-end manner. This is achieved by using a Connectionist Temporal Classification (CTC) loss to bind the recognition and translation problems into a single unified architecture. This joint approach does not require any ground-truth timing information, simultaneously solving two co-dependant sequence-to-sequence learning problems and leads to significant performance gains. We evaluate the recognition and translation performances of our approaches on the challenging RWTH-PHOENIX-Weather-2014T (PHOENIX14T) dataset. We report state-of-the-art sign language recognition and translation results achieved by our Sign Language Transformers. Our translation networks outperform both sign video to spoken language and gloss to spoken language translation models, in some cases more than doubling the performance (9.58 vs. 21.80 BLEU-4 Score). We also share new baseline translation results using transformer networks for several other text-to-text sign language translation tasks.

Related Papers

• → Neural Sign Language Translation(2018)682 cited
• → Continuous sign language recognition: Towards large vocabulary statistical recognition systems handling multiple signers(2015)494 cited
• → Video-Based Sign Language Recognition Without Temporal Segmentation(2018)412 cited
• → SubUNets: End-to-End Hand Shape and Continuous Sign Language Recognition(2017)361 cited