Breaking the Silence: A Dataset and Benchmark for Bangla Text-to-Gloss Translation
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866917453482688512 |
|---|---|
| author | Abdullah, Sharif Mohammad Paul, Abhijit Dipta, Shubhashis Roy Masud, Zarif Rayana, Shebuti Kabir, Ahmedul |
| author_facet | Abdullah, Sharif Mohammad Paul, Abhijit Dipta, Shubhashis Roy Masud, Zarif Rayana, Shebuti Kabir, Ahmedul |
| contents | Gloss is a written approximation that bridges Sign Language (SL) and its corresponding spoken language. Despite a deaf and hard-of-hearing population of at least 3 million in Bangladesh, Bangla Sign Language (BdSL) remains largely understudied, with no prior work on Bangla text-to-gloss translation and no publicly available datasets. To address this gap, we construct the first Bangla text-to-gloss dataset, consisting of 1,000 manually annotated and 4,000 synthetically generated Bangla sentence-gloss pairs, along with 159 expert human-annotated pairs used as a test set. Our experimental framework performs a comparative analysis between several fine-tuned open-source models and a leading closed-source LLM to evaluate their performance in low-resource BdSL translation. GPT-5.4 achieves the best overall performance, while a fine-tuned mBART model performs competitively despite being approximately 100% smaller. Qwen-3 outperforms all other models in human evaluation. This work introduces the first dataset and trained model for Bangla text-to-gloss translation. It also demonstrates the effectiveness of systematically generated synthetic data for addressing challenges in low-resource sign language translation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_02293 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Breaking the Silence: A Dataset and Benchmark for Bangla Text-to-Gloss Translation Abdullah, Sharif Mohammad Paul, Abhijit Dipta, Shubhashis Roy Masud, Zarif Rayana, Shebuti Kabir, Ahmedul Computation and Language Artificial Intelligence Gloss is a written approximation that bridges Sign Language (SL) and its corresponding spoken language. Despite a deaf and hard-of-hearing population of at least 3 million in Bangladesh, Bangla Sign Language (BdSL) remains largely understudied, with no prior work on Bangla text-to-gloss translation and no publicly available datasets. To address this gap, we construct the first Bangla text-to-gloss dataset, consisting of 1,000 manually annotated and 4,000 synthetically generated Bangla sentence-gloss pairs, along with 159 expert human-annotated pairs used as a test set. Our experimental framework performs a comparative analysis between several fine-tuned open-source models and a leading closed-source LLM to evaluate their performance in low-resource BdSL translation. GPT-5.4 achieves the best overall performance, while a fine-tuned mBART model performs competitively despite being approximately 100% smaller. Qwen-3 outperforms all other models in human evaluation. This work introduces the first dataset and trained model for Bangla text-to-gloss translation. It also demonstrates the effectiveness of systematically generated synthetic data for addressing challenges in low-resource sign language translation. |
| title | Breaking the Silence: A Dataset and Benchmark for Bangla Text-to-Gloss Translation |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2504.02293 |