Sign language production (SLP) is the process of generating sign language videos from spoken language expressions. Since sign languages are highly under-resourced, existing vision-based SLP approaches suffer from out-of-vocabulary (OOV) and test-time generalization problems and thus generate low-quality translations. To address these problems, we introduce an avatar-based SLP system composed of a sign language translation (SLT) model and an avatar animation generation module. Our Transformer-based SLT model utilizes two additional strategies to resolve these problems: named entity transformation to reduce OOV tokens and context vector generation using a pretrained language model (e.g., BERT) to reliably train the decoder. Our system is validated on a new Korean-Korean Sign Language (KSL) dataset of weather forecasts and emergency announcements. Our SLT model achieves an 8.77 higher BLEU-4 score and a 4.57 higher ROUGE-L score over those of our baseline model. In a user evaluation, 93.48% of named entities were successfully identified by participants, demonstrating marked improvement on OOV issues.

We introduce a new sign language production (SLP) and sign language translation (SLT) dataset, NIASL2021, consisting of 201,026 Korean-KSL data pairs. KSL translations of Korean source texts are represented in three formats: video recordings, keypoint position data, and time-aligned gloss annotations for each hand (using a 7,989 sign vocabulary) and for eight different non-manual signals (NMS). We evaluated our sign language elicitation methodology and found that text-based prompting had a negative effect on translation quality in terms of naturalness and comprehension. We recommend distilling text into a visual medium before translating into sign language or adding a prompt-blind review step to text-based translation methodologies.