MELLA: Bridging Linguistic Capability and Cultural Groundedness for Low-Resource Language MLLMs

Gao, Yufei; Fei, Jiaying; Chen, Nuo; Chen, Ruirui; Yan, Guohang; Lan, Yunshi; Shi, Botian

doi:10.24963/ijcai.2026/125

MELLA: Bridging Linguistic Capability and Cultural Groundedness for Low-Resource Language MLLMs

Yufei Gao, Jiaying Fei, Nuo Chen, Ruirui Chen, Guohang Yan, Yunshi Lan, Botian Shi

Download BibTeX Cite

Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence. Main Track. Pages 1115-1123.

https://doi.org/10.24963/ijcai.2026/125

Keywords

Computer Vision: Vision, language and reasoning

Abstract

Multimodal Large Language Models (MLLMs) perform strongly in high-resource languages, yet often produce fluent but culturally "thin" descriptions in low-resource settings. We argue that this failure is not merely a linguistic limitation: culture-specific visual knowledge depends on native visual-textual alignments that translation-centric pipelines rarely provide.

We present MELLA, a multimodal dataset across eight low-resource languages, designed to jointly support linguistic fluency and cultural groundedness. MELLA uses a dual-source strategy that combines native web image-alt-text pairs for culture-grounded supervision with generated-and-translated image descriptions for linguistically rich supervision, explicitly separating two learning signals often conflated in multilingual multimodal data.

Through controlled diagnostic fine-tuning on multiple MLLM backbones, we show that MELLA mitigates cultural hallucination by helping models recognize and articulate culturally specific entities overlooked by translation-based adaptation. Our findings highlight data alignment, rather than model modification alone, as a key path toward culturally grounded multimodal understanding in low-resource languages.

Yufei Gao, Jiaying Fei, Nuo Chen, Ruirui Chen, Guohang Yan, Yunshi Lan, and Botian Shi. 2026. MELLA: Bridging Linguistic Capability and Cultural Groundedness for Low-Resource Language MLLMs. In Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence, IJCAI-26. IJCAI, 1115–1123. https://doi.org/10.24963/ijcai.2026/125

@inproceedings{ijcai2026p125,
  title     = {MELLA: Bridging Linguistic Capability and Cultural Groundedness for Low-Resource Language MLLMs},
  author    = {Gao, Yufei and Fei, Jiaying and Chen, Nuo and Chen, Ruirui and Yan, Guohang and Lan, Yunshi and Shi, Botian},
  booktitle = {Proceedings of the Thirty-Fifth International Joint Conference on
               Artificial Intelligence, {IJCAI-26}},
  publisher = {International Joint Conferences on Artificial Intelligence Organization},
  editor    = {Diego Calvanese},
  pages     = {1115--1123},
  year      = {2026},
  month     = {8},
  note      = {Main Track},
  doi       = {10.24963/ijcai.2026/125},
  url       = {https://doi.org/10.24963/ijcai.2026/125},
}

TY  - CPAPER
TI  - MELLA: Bridging Linguistic Capability and Cultural Groundedness for Low-Resource Language MLLMs
AU  - Gao, Yufei
AU  - Fei, Jiaying
AU  - Chen, Nuo
AU  - Chen, Ruirui
AU  - Yan, Guohang
AU  - Lan, Yunshi
AU  - Shi, Botian
T2  - Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence, IJCAI-26
PB  - International Joint Conferences on Artificial Intelligence Organization
PY  - 2026
SP  - 1115
EP  - 1123
DO  - 10.24963/ijcai.2026/125
UR  - https://doi.org/10.24963/ijcai.2026/125
ER  -