---
title: "Crowdsource, Crawl, or Generate? Creating SEA-VL, a Multicultural Vision-Language Dataset for Southeast Asia"
canonical_url: "https://www.modelscope.cn/papers/125634"
md_url: "https://www.modelscope.cn/papers/125634.md"
arxiv_id: 2503.07920
published: 2025-03-18
last_updated: 2025-03-18
authors:
  - "Samuel Cahyawijaya"
  - "Holy Lovenia"
  - "Joel Ruben Antony Moniz"
  - "Tack Hwa Wong"
  - "Mohammad Rifqi Farhansyah"
  - "Thant Thiri Maung"
  - "Frederikus Hudi"
  - "David Anugraha"
  - "Muhammad Ravi Shulthan Habibi"
  - "Muhammad Reza Qorib"
  - "Amit Agarwal"
  - "Joseph Marvin Imperial"
  - "Hitesh Laxmichand Patel"
  - "Vicky Feliren"
  - "Bahrul Ilmi Nasution"
  - "Manuel Antonio Rufino"
  - "Genta Indra Winata"
  - "Rian Adam Rajagede"
  - "Carlos Rafael Catalan"
  - "Mohamed Fazli Imam"
  - "Priyaranjan Pattnayak"
  - "Salsabila Zahirah Pranida"
  - "Kevin Pratama"
  - "Yeshil Bangera"
  - "Adisai Na-Thalang"
  - "Patricia Nicole Monderin"
  - "Yueqi Song"
  - "Christian Simon"
  - "Lynnette Hui Xian Ng"
  - "Richardy Lobo' Sapan"
  - "Taki Hasan Rafi"
  - "Bin Wang"
  - Supryadi
  - "Kanyakorn Veerakanjana"
  - "Piyalitt Ittichaiwong"
  - "Matthew Theodore Roque"
  - "Karissa Vincentio"
  - "Takdanai Kreangphet"
  - "Phakphum Artkaew"
  - "Kadek Hendrawan Palgunadi"
  - "Yanzhi Yu"
  - "Rochana Prih Hastuti"
  - "William Nixon"
  - "Mithil Bangera"
  - "Adrian Xuan Wei Lim"
  - "Aye Hninn Khine"
  - "Hanif Muhammad Zhafran"
  - "Teddy Ferdinan"
  - "Audra Aurora Izzani"
  - "Ayushman Singh"
  - Evan
  - "Jauza Akbar Krito"
  - "Michael Anugraha"
  - "Fenal Ashokbhai Ilasariya"
  - "Haochen Li"
  - "John Amadeo Daniswara"
  - "Filbert Aurelian Tjiaranata"
  - "Eryawan Presma Yulianrifat"
  - "Can Udomcharoenchaikit"
  - "Fadil Risdian Ansori"
  - "Mahardika Krisna Ihsani"
  - "Giang Nguyen"
  - "Anab Maulana Barik"
  - "Dan John Velasco"
  - "Rifo Ahmad Genadi"
  - "Saptarshi Saha"
  - "Chengwei Wei"
  - "Isaiah Flores"
  - "Kenneth Ko Han Chen"
  - "Anjela Gail Santos"
  - "Wan Shen Lim"
  - "Kaung Si Phyo"
  - "Tim Santos"
  - "Meisyarah Dwiastuti"
  - "Jiayun Luo"
  - "Jan Christian Blaise Cruz"
  - "Ming Shan Hee"
  - "Ikhlasul Akmal Hanif"
  - "M. Alif Al Hakim"
  - "Muhammad Rizky Sya'ban"
  - "Kun Kerdthaisong"
  - "Lester James V. Miranda"
  - "Fajri Koto"
  - "Tirana Noor Fatyanosa"
  - "Alham Fikri Aji"
  - "Jostin Jerico Rosal"
  - "Jun Kevin"
  - "Robert Wijaya"
  - "Onno P. Kampman"
  - "Ruochen Zhang"
  - "Börje F. Karlsson"
  - "Peerat Limkonchotiwat"
model_name: SEA-VL
model_developer: "Cohere、SEACrowd、IndoNLP、Mila - Quebec AI Institute、Polytechnique Montreal、Bandung Institute of Technology、Ton Duc Thang University、Nara Institute of Science and Technology、University of Toronto、University of Indonesia、National University of Singapore、Oracle、University of Bath、National University Philippines、Monash University, Indonesia、The University of Manchester、Samsung R&D Institute Philippines、Capital One、Universitas Islam Indonesia、MBZUAI、Meta、University of New Haven、SCB 10X、Carnegie Mellon University、Sony Group Corporation、Institute for Infocomm Research, Singapore、Tianjin University、Mahidol University、Binus University、Srinakharinwirot University、New York University、Institut Teknologi Sepuluh Nopember、Macau University of Science and Technology、Universitas Gadjah Mada、King Mongkut's University of Technology Thonburi、Wrocław Tech、University of Illinois Urbana-Champaign、Vidyasirimedhi Institute of Science and Technology、Auburn University、Indian Statistical Institute, Kolkata、Ateneo de Manila University、Singapore Polytechnic、University of the Philippines、Graphcore、Dataxet:Sonar、Singapore University of Technology and Design、Thammasat University、Allen AI、Brawijaya University、Seoul National University of Science and Technology、Universitas Pelita Harapan、MOH Office for Healthcare Transformation、Brown University、Beijing Academy of Artificial Intelligence (BAAI)、AI Singapore、Chulalongkorn University"
domain:
  - "计算机视觉"
  - "自然语言处理"
  - "多模态学习"
  - "数据集构建"
  - "低资源语言"
type:
  - "计算机视觉"
  - "自然语言处理"
  - "多模态学习"
  - "数据集构建"
  - "低资源语言"
  - "Computer Vision and Pattern Recognition"
  - "Artificial Intelligence"
  - "Computation and Language"
arxiv_url: "https://arxiv.org/abs/2503.07920"
pdf_url: "https://arxiv.org/pdf/2503.07920"
code_link: "https://github.com/SEACrowd/sea-vl-experiments"
---

# Crowdsource, Crawl, or Generate? Creating SEA-VL, a Multicultural Vision-Language Dataset for Southeast Asia

> Southeast Asia (SEA) is a region of extraordinary linguistic and cultural diversity, yet it remains significantly underrepresented in vision-language (VL) research. This often results in artificial intelligence (AI) models that fail to capture SEA cultural…

「Crowdsource, Crawl, or Generate? Creating SEA-VL, a Multicultural Vision-Language Dataset for Southeast Asia」是 ModelScope 魔搭社区收录的论文，arXiv 2503.07920，作者为 Samuel Cahyawijaya, Holy Lovenia, Joel Ruben Antony Moniz et al.，发表于 2025-03-18，属于 计算机视觉、自然语言处理、多模态学习 领域。

- **ArXiv**: 2503.07920
- **Published**: 2025-03-18
- **Authors**: Samuel Cahyawijaya, Holy Lovenia, Joel Ruben Antony Moniz, Tack Hwa Wong, Mohammad Rifqi Farhansyah, Thant Thiri Maung, Frederikus Hudi, David Anugraha, Muhammad Ravi Shulthan Habibi, Muhammad Reza Qorib, Amit Agarwal, Joseph Marvin Imperial, Hitesh Laxmichand Patel, Vicky Feliren, Bahrul Ilmi Nasution, Manuel Antonio Rufino, Genta Indra Winata, Rian Adam Rajagede, Carlos Rafael Catalan, Mohamed Fazli Imam, Priyaranjan Pattnayak, Salsabila Zahirah Pranida, Kevin Pratama, Yeshil Bangera, Adisai Na-Thalang, Patricia Nicole Monderin, Yueqi Song, Christian Simon, Lynnette Hui Xian Ng, Richardy Lobo' Sapan, Taki Hasan Rafi, Bin Wang, Supryadi, Kanyakorn Veerakanjana, Piyalitt Ittichaiwong, Matthew Theodore Roque, Karissa Vincentio, Takdanai Kreangphet, Phakphum Artkaew, Kadek Hendrawan Palgunadi, Yanzhi Yu, Rochana Prih Hastuti, William Nixon, Mithil Bangera, Adrian Xuan Wei Lim, Aye Hninn Khine, Hanif Muhammad Zhafran, Teddy Ferdinan, Audra Aurora Izzani, Ayushman Singh, Evan, Jauza Akbar Krito, Michael Anugraha, Fenal Ashokbhai Ilasariya, Haochen Li, John Amadeo Daniswara, Filbert Aurelian Tjiaranata, Eryawan Presma Yulianrifat, Can Udomcharoenchaikit, Fadil Risdian Ansori, Mahardika Krisna Ihsani, Giang Nguyen, Anab Maulana Barik, Dan John Velasco, Rifo Ahmad Genadi, Saptarshi Saha, Chengwei Wei, Isaiah Flores, Kenneth Ko Han Chen, Anjela Gail Santos, Wan Shen Lim, Kaung Si Phyo, Tim Santos, Meisyarah Dwiastuti, Jiayun Luo, Jan Christian Blaise Cruz, Ming Shan Hee, Ikhlasul Akmal Hanif, M. Alif Al Hakim, Muhammad Rizky Sya'ban, Kun Kerdthaisong, Lester James V. Miranda, Fajri Koto, Tirana Noor Fatyanosa, Alham Fikri Aji, Jostin Jerico Rosal, Jun Kevin, Robert Wijaya, Onno P. Kampman, Ruochen Zhang, Börje F. Karlsson, Peerat Limkonchotiwat
- **Model**: SEA-VL
- **Developer**: Cohere、SEACrowd、IndoNLP、Mila - Quebec AI Institute、Polytechnique Montreal、Bandung Institute of Technology、Ton Duc Thang University、Nara Institute of Science and Technology、University of Toronto、University of Indonesia、National University of Singapore、Oracle、University of Bath、National University Philippines、Monash University, Indonesia、The University of Manchester、Samsung R&D Institute Philippines、Capital One、Universitas Islam Indonesia、MBZUAI、Meta、University of New Haven、SCB 10X、Carnegie Mellon University、Sony Group Corporation、Institute for Infocomm Research, Singapore、Tianjin University、Mahidol University、Binus University、Srinakharinwirot University、New York University、Institut Teknologi Sepuluh Nopember、Macau University of Science and Technology、Universitas Gadjah Mada、King Mongkut's University of Technology Thonburi、Wrocław Tech、University of Illinois Urbana-Champaign、Vidyasirimedhi Institute of Science and Technology、Auburn University、Indian Statistical Institute, Kolkata、Ateneo de Manila University、Singapore Polytechnic、University of the Philippines、Graphcore、Dataxet:Sonar、Singapore University of Technology and Design、Thammasat University、Allen AI、Brawijaya University、Seoul National University of Science and Technology、Universitas Pelita Harapan、MOH Office for Healthcare Transformation、Brown University、Beijing Academy of Artificial Intelligence (BAAI)、AI Singapore、Chulalongkorn University
- **Domain**: 计算机视觉, 自然语言处理, 多模态学习, 数据集构建, 低资源语言
- **ArXiv URL**: https://arxiv.org/abs/2503.07920
- **PDF**: https://arxiv.org/pdf/2503.07920
- **Code**: https://github.com/SEACrowd/sea-vl-experiments

Source: https://www.modelscope.cn/papers/125634

---

> 众包、爬取还是生成？创建 SEA-VL：面向东南亚的多元文化视觉-语言数据集

## 摘要

SEA-VL 是一个面向东南亚（SEA）地区的开源多元文化视觉-语言（VL）数据集，旨在解决东南亚1300多种语言及其文化在AI数据中严重代表性不足的问题。该数据集包含约128万张具有文化相关性的图像，规模超过现有同类数据集50倍以上。项目通过众包、网络爬取与过滤、以及合成生成三种策略收集数据，并系统评估了各方法在文化相关性、效率和质量上的权衡。结果表明，网络爬取结合语义相似度过滤是大规模构建训练集的最优方案，而当前扩散模型生成的合成图像在准确反映东南亚文化细节方面仍不可靠。

## Abstract

Southeast Asia (SEA) is a region of extraordinary linguistic and cultural diversity, yet it remains significantly underrepresented in vision-language (VL) research. This often results in artificial intelligence (AI) models that fail to capture SEA cultural nuances. To fill this gap, we present SEA-VL, an open-source initiative dedicated to developing high-quality, culturally relevant data for SEA languages. By involving contributors from SEA countries, SEA-VL aims to ensure better cultural relevance and diversity, fostering greater inclusivity of underrepresented languages in VL research. Beyond crowdsourcing, our initiative goes one step further in the exploration of the automatic collection of culturally relevant images through crawling and image generation. First, we find that image crawling achieves approximately ~85% cultural relevance while being more cost- and time-efficient than crowdsourcing. Second, despite the substantial progress in generative vision models, synthetic images remain unreliable in accurately reflecting SEA cultures. The generated images often fail to reflect the nuanced traditions and cultural contexts of the region. Collectively, we gather 1.28M SEA culturally-relevant images, more than 50 times larger than other existing datasets. Through SEA-VL, we aim to bridge the representation gap in SEA, fostering the development of more inclusive AI systems that authentically represent diverse cultures across SEA.
