FTibSuite: A Comprehensive Resource Suite for Tibetan Vision-Language Modeling
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866916049180426240 |
|---|---|
| author | Xu, Guixian Liang, Yide Su, Zeli Song, Xuexian Zhang, Ziyin Dong, Yushuang Zhang, Ting Han, Xu |
| author_facet | Xu, Guixian Liang, Yide Su, Zeli Song, Xuexian Zhang, Ziyin Dong, Yushuang Zhang, Ting Han, Xu |
| contents | Vision-language models have progressed rapidly, but Tibetan remains a severely underserved low-resource language due to the lack of reproducible training and evaluation infrastructure. To fill this gap, we introduce FTibSuite, a comprehensive resource suite for Tibetan vision-language research, consisting of FTibData (human-verified multimodal training corpora spanning continual pretraining, image-text alignment, and instruction tuning data), FTibBench (Tibetan adaptations of five mainstream multimodal benchmarks with a hierarchical quality-control workflow to reduce translation noise), and FTibVLM, a reproducible baseline built on Qwen3-VL-8B-Instruct via a three-stage adaptation pipeline. Experiments on FTibBench show FTibVLM delivers consistent performance gains across all tasks, such as improving MMBench accuracy from 42.97 to 67.78 and POPE-random accuracy from 47.53 to 80.56, while retaining the backbone's original Chinese capabilities with minimal degradation, providing the first standardized foundation for Tibetan multimodal research. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_26601 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | FTibSuite: A Comprehensive Resource Suite for Tibetan Vision-Language Modeling Xu, Guixian Liang, Yide Su, Zeli Song, Xuexian Zhang, Ziyin Dong, Yushuang Zhang, Ting Han, Xu Computer Vision and Pattern Recognition Vision-language models have progressed rapidly, but Tibetan remains a severely underserved low-resource language due to the lack of reproducible training and evaluation infrastructure. To fill this gap, we introduce FTibSuite, a comprehensive resource suite for Tibetan vision-language research, consisting of FTibData (human-verified multimodal training corpora spanning continual pretraining, image-text alignment, and instruction tuning data), FTibBench (Tibetan adaptations of five mainstream multimodal benchmarks with a hierarchical quality-control workflow to reduce translation noise), and FTibVLM, a reproducible baseline built on Qwen3-VL-8B-Instruct via a three-stage adaptation pipeline. Experiments on FTibBench show FTibVLM delivers consistent performance gains across all tasks, such as improving MMBench accuracy from 42.97 to 67.78 and POPE-random accuracy from 47.53 to 80.56, while retaining the backbone's original Chinese capabilities with minimal degradation, providing the first standardized foundation for Tibetan multimodal research. |
| title | FTibSuite: A Comprehensive Resource Suite for Tibetan Vision-Language Modeling |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2605.26601 |