Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.11490 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911600378642432 |
|---|---|
| author | Cahyawijaya, Samuel Limkonchotiwat, Peerat Wong, Tack Hwa Patel, Hitesh Laxmichand Agarwal, Amit Rufino, Manuel Antonio Catalan, Carlos Rafael Qorib, Muhammad Reza Feliren, Vicky Lovenia, Holy Khine, Aye Hninn Hudi, Frederikus Anugraha, David Aji, Alham Fikri Chumpu, Romrawin Pham, Viet-Thanh Wang, Minghan Imam, Mohamed Fazli Zhang, Ruochen Imperial, Joseph Marvin Nur'aini, Khumaisa Long, Do Xuan Wijanarko, Musa Izzanardi Moniz, Joel Ruben Antony Irawan, Patrick Amadeus Zhafran, Hanif Muhammad Flores, Isaiah Pranida, Salsabila Zahirah Kevin, Jun Rosal, Jostin Jerico Monderin, Patricia Nicole Kerdthaisong, Kun Mustafid, Ahmad Nguyen, My Chiffon Jongwiriyanurak, Natchapon Worajitwannakul, Siva Li, Haochen Lim, Adrian Xuan Wei Wang, Bin Habibi, Muhammad Ravi Shulthan Ng, Lynnette Hui Xian Bangera, Mithil Bangera, Yeshil Pattnayak, Priyaranjan Chan, Dun Li Djuniwar, Sherissa Caren Oo, Cho Chan Myei Shan, Hee Ming |
| author_facet | Cahyawijaya, Samuel Limkonchotiwat, Peerat Wong, Tack Hwa Patel, Hitesh Laxmichand Agarwal, Amit Rufino, Manuel Antonio Catalan, Carlos Rafael Qorib, Muhammad Reza Feliren, Vicky Lovenia, Holy Khine, Aye Hninn Hudi, Frederikus Anugraha, David Aji, Alham Fikri Chumpu, Romrawin Pham, Viet-Thanh Wang, Minghan Imam, Mohamed Fazli Zhang, Ruochen Imperial, Joseph Marvin Nur'aini, Khumaisa Long, Do Xuan Wijanarko, Musa Izzanardi Moniz, Joel Ruben Antony Irawan, Patrick Amadeus Zhafran, Hanif Muhammad Flores, Isaiah Pranida, Salsabila Zahirah Kevin, Jun Rosal, Jostin Jerico Monderin, Patricia Nicole Kerdthaisong, Kun Mustafid, Ahmad Nguyen, My Chiffon Jongwiriyanurak, Natchapon Worajitwannakul, Siva Li, Haochen Lim, Adrian Xuan Wei Wang, Bin Habibi, Muhammad Ravi Shulthan Ng, Lynnette Hui Xian Bangera, Mithil Bangera, Yeshil Pattnayak, Priyaranjan Chan, Dun Li Djuniwar, Sherissa Caren Oo, Cho Chan Myei Shan, Hee Ming |
| contents | While the field of vision-language (VL) has achieved remarkable success in integrating visual and textual information across multiple languages and domains, there is still no dedicated framework for assessing human-centric alignment in vision-language systems. We offer two contributions to address this gap. First, we introduce Anthropogenic Regional Adaptation: a novel paradigm that aims to optimize model relevance to specific regional contexts while ensuring the retention of global generalization capabilities. Second, we present a simple, but effective adaptation method named Geographical-generalization-made-easy (GG-EZ), which utilizes regional data filtering and model merging. Through comprehensive experiments on 3 VL architectures: large vision-language models, text-to-image diffusion models, and vision-language embedding models, and a case study in Southeast Asia (SEA) regional adaptation, we demonstrate the importance of Anthropogenic Regional Adaptation and the effectiveness of GG-EZ, showing 5-15% gains in cultural relevance metrics across SEA while maintaining over 98% of global performance and even occasionally surpassing it. Our findings establish Anthropogenic Regional Alignment as a foundational paradigm towards applicability of multimodal vision-language models in diverse regions and demonstrate a simple-yet-effective baseline method that optimizes regional value alignment while preserving global generalization. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_11490 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Anthropogenic Regional Adaptation in Multimodal Vision-Language Model Cahyawijaya, Samuel Limkonchotiwat, Peerat Wong, Tack Hwa Patel, Hitesh Laxmichand Agarwal, Amit Rufino, Manuel Antonio Catalan, Carlos Rafael Qorib, Muhammad Reza Feliren, Vicky Lovenia, Holy Khine, Aye Hninn Hudi, Frederikus Anugraha, David Aji, Alham Fikri Chumpu, Romrawin Pham, Viet-Thanh Wang, Minghan Imam, Mohamed Fazli Zhang, Ruochen Imperial, Joseph Marvin Nur'aini, Khumaisa Long, Do Xuan Wijanarko, Musa Izzanardi Moniz, Joel Ruben Antony Irawan, Patrick Amadeus Zhafran, Hanif Muhammad Flores, Isaiah Pranida, Salsabila Zahirah Kevin, Jun Rosal, Jostin Jerico Monderin, Patricia Nicole Kerdthaisong, Kun Mustafid, Ahmad Nguyen, My Chiffon Jongwiriyanurak, Natchapon Worajitwannakul, Siva Li, Haochen Lim, Adrian Xuan Wei Wang, Bin Habibi, Muhammad Ravi Shulthan Ng, Lynnette Hui Xian Bangera, Mithil Bangera, Yeshil Pattnayak, Priyaranjan Chan, Dun Li Djuniwar, Sherissa Caren Oo, Cho Chan Myei Shan, Hee Ming Artificial Intelligence Computation and Language Computer Vision and Pattern Recognition While the field of vision-language (VL) has achieved remarkable success in integrating visual and textual information across multiple languages and domains, there is still no dedicated framework for assessing human-centric alignment in vision-language systems. We offer two contributions to address this gap. First, we introduce Anthropogenic Regional Adaptation: a novel paradigm that aims to optimize model relevance to specific regional contexts while ensuring the retention of global generalization capabilities. Second, we present a simple, but effective adaptation method named Geographical-generalization-made-easy (GG-EZ), which utilizes regional data filtering and model merging. Through comprehensive experiments on 3 VL architectures: large vision-language models, text-to-image diffusion models, and vision-language embedding models, and a case study in Southeast Asia (SEA) regional adaptation, we demonstrate the importance of Anthropogenic Regional Adaptation and the effectiveness of GG-EZ, showing 5-15% gains in cultural relevance metrics across SEA while maintaining over 98% of global performance and even occasionally surpassing it. Our findings establish Anthropogenic Regional Alignment as a foundational paradigm towards applicability of multimodal vision-language models in diverse regions and demonstrate a simple-yet-effective baseline method that optimizes regional value alignment while preserving global generalization. |
| title | Anthropogenic Regional Adaptation in Multimodal Vision-Language Model |
| topic | Artificial Intelligence Computation and Language Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2604.11490 |