arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2609.30739cs.CL

SEA-CLIP-Tiny:面向东南亚语言的高效多语言文本-视觉嵌入模型

SEA-CLIP-Tiny: Efficient Multilingual Text-Vision Embedding for Southeast Asian Languages

Puja Ahmad Habibi, Faiz Assabil Firdaus, Ashvanth S, Ekapol Chuangsuwanich, Pume Tuchinda, Peerat Limkonchotiwat

首次发表
浏览论文内容

中文总结 AI 辅助

本文提出SEA-CLIP-Tiny,一个参数少于50M的紧凑多语言文本-视觉嵌入模型,通过区域数据整理和CLIP-KD风格框架适配东南亚语言,在七种语言上取得最优平均检索性能,并比MobileCLIP2参数减少38.4%、R@10提升12.1点。

中文摘要 AI 辅助

多语言文本-视觉嵌入模型对于跨语言图像-文本检索至关重要,但由于东南亚地区的语言多样性以及有限的数据和计算资源,该地区的语言在现有模型中支持不足。在本文中,我们介绍了SEA-CLIP-Tiny,一个面向东南亚的紧凑型多语言文本-视觉嵌入模型,其参数少于5000万。我们的模型通过区域数据整理和多语言教师指导,将CLIP-KD风格的框架适配到东南亚多语言设置中。在七种东南亚语言上的实验表明,SEA-CLIP-Tiny在评估的学生模型中取得了最强的平均检索性能,在R@1、R@5和R@10上分别达到12.9%、31.5%和42.2%。与MobileCLIP2相比,它在使用参数减少38.4%且CPU延迟更低的情况下,将平均R@10提高了12.1个百分点。这些结果凸显了区域感知训练对于东南亚高效多语言文本-视觉模型的重要性。

英文摘要

Multilingual text-vision embedding models are essential for cross-lingual image-text retrieval, but Southeast Asian languages remain poorly supported due to the region's linguistic diversity and limited data and computing resources. In this paper, we introduce SEA-CLIP-Tiny, a compact multilingual text-vision embedding model for Southeast Asia with fewer than 50M parameters. Our model adapts a CLIP-KD-style framework to Southeast Asian multilingual settings through regional data curation and multilingual teacher guidance. Experiments across seven Southeast Asian languages show that SEA-CLIP-Tiny achieves the strongest average retrieval performance among the evaluated student models, reaching 12.9%, 31.5%, and 42.2% at R@1, R@5, and R@10, respectively. Compared with MobileCLIP2, it improves average R@10 by 12.1 points while using 38.4% fewer parameters and lower measured CPU latency. These results highlight the importance of region-aware training for efficient multilingual text-vision models in Southeast Asia.

发表机构

  • SEACrowd
  • University of Indonesia(印度尼西亚大学)
  • Cohere Labs Community(Cohere Labs 社区)
  • Technical University of Denmark(丹麦技术大学)
  • Chulalongkorn University(朱拉隆功大学)
  • Vidyasirimedhi Institute of Science and Technology(维迪亚西里梅迪科学技术研究所)
  • AI Singapore(新加坡人工智能)

机构由 AI 辅助整理,请以论文原文为准。

补充信息

↑