BackTranslation2.0: A Linguistically Motivated Metric to Assess Sign Language Production
Cory, O., Ivashechkin, M., Ranum, O., Low, J., Fish, E., Pelykh, A., Sahin, K., Mercanoglu Sincan, O., Bowden, R. (2026). "BackTranslation2.0: A Linguistically Motivated Metric to Assess Sign Language Production." ECCV 2026. https://arxiv.org/pdf/2606.28673
We introduce BackTranslation2.0, a linguistically grounded evaluation metric for text-to-sign translation that moves beyond naïve backtranslation. The approach adopts an agentic framework in which a deterministic pipeline orchestrates a suite of specialised tools to assess four scoring dimensions (grammatical correctness, phonological accuracy, motion fluency, and generation fidelity), with LLM-based cross-referential comparison modules checking consistency across tools against linguistic expectations. Validated on a British Sign Language dataset annotated by native Deaf raters, the metric demonstrates stronger correlation with human judgements than six baseline metrics.