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

arXiv: https://arxiv.org/abs/2609.30739 · HTML: https://mlsanity.com/paper/2609.30739/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 24 of its 28 references that are in the map.
- First posted on arXiv 25 Sep 2026.

## Builds on (most cited first, 24 of 24)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks](https://mlsanity.com/paper/2312.14238/) (2023)
- [Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond](https://mlsanity.com/paper/2308.12966/) (2023)
- [Conceptual 12M: Pushing Web-Scale Image-Text Pre-Training To Recognize Long-Tail Visual Concepts](https://mlsanity.com/paper/2102.08981/) (2021)
- [SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features](https://mlsanity.com/paper/2502.14786/) (2025)
- [WIT: Wikipedia-based Image Text Dataset for Multimodal Multilingual Machine Learning](https://mlsanity.com/paper/2103.01913/) (2021)
- [Data Filtering Networks](https://mlsanity.com/paper/2309.17425/) (2023)
- [VLM2Vec: Training Vision-Language Models for Massive Multimodal Embedding Tasks](https://mlsanity.com/paper/2410.05160/) (2024)
- [Chinese CLIP: Contrastive Vision-Language Pretraining in Chinese](https://mlsanity.com/paper/2211.01335/) (2022)
- [Crossmodal-3600: A Massively Multilingual Multimodal Evaluation Dataset](https://mlsanity.com/paper/2205.12522/) (2022)
- [TinyCLIP: CLIP Distillation via Affinity Mimicking and Weight Inheritance](https://mlsanity.com/paper/2309.12314/) (2023)
- [CLIP-KD: An Empirical Study of CLIP Model Distillation](https://mlsanity.com/paper/2307.12732/) (2023)
- [AltCLIP: Altering the Language Encoder in CLIP for Extended Language Capabilities](https://mlsanity.com/paper/2211.06679/) (2022)
- [Meta CLIP 2: A Worldwide Scaling Recipe](https://mlsanity.com/paper/2507.22062/) (2025)
- [All Languages Matter: Evaluating LMMs on Culturally Diverse 100 Languages](https://mlsanity.com/paper/2411.16508/) (2024)
- [Bloom Library: Multimodal Datasets in 300+ Languages for a Variety of Downstream Tasks](https://mlsanity.com/paper/2210.14712/) (2022)
- [MobileCLIP2: Improving Multi-Modal Reinforced Training](https://mlsanity.com/paper/2508.20691/) (2025)
- [Contrastive Language-Image Pre-training for the Italian Language](https://mlsanity.com/paper/2108.08688/) (2021)
- [SEA-HELM: Southeast Asian Holistic Evaluation of Language Models](https://mlsanity.com/paper/2502.14301/) (2025)
- [Grounding Multilingual Multimodal LLMs With Cultural Knowledge](https://mlsanity.com/paper/2508.07414/) (2025)
- [MIEB: Massive Image Embedding Benchmark](https://mlsanity.com/paper/2504.10471/) (2025)
- [Babel-ImageNet: Massively Multilingual Evaluation of Vision-and-Language Representations](https://mlsanity.com/paper/2306.08658/) (2023)
- [SEA-Guard: Culturally Grounded Multilingual Safeguard for Southeast Asia](https://mlsanity.com/paper/2602.01618/) (2026)
- [Anthropogenic Regional Adaptation in Multimodal Vision-Language Model](https://mlsanity.com/paper/2604.11490/) (2026)

## Abstract (arXiv)

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.