MICo-150K is a large-scale synthetic dataset generated by Nano Banana and Nano Banana Pro, designed to advance open-source models in Multi-Image Composition (MICo).
We fine-tune a diverse set of base models—including Qwen-Image, BAGEL, OmniGen2, Lumina-DiMOO, and BLIP3o-Next-Edit, on MICo-150K. All models demonstrate substantial performance improvements on our proposed MICo Bench after fine-tuning. Notably, Qwen-Image, originally developed as a text-to-image model, achieves remarkable gains after adaptation. The fine-tuned variant, Qwen-Image-MICo, surpasses Qwen-Image-Edit-2509 on both MICo Bench and OmniContext benchmarks, highlighting its strong generalization capability and broad applicability in multi-image composition scenarios.

We organize the MICo-150K dataset into three primary categories and one specialized task subset, De&Re. Each primary category encompasses multiple sub-tasks:
The De&Re subset focuses on decomposition and recomposition tasks. Specifically, elements from a real-world image (e.g., persons, apparel, scenes) are first decomposed into multiple component images. These components are subsequently recomposed into a single composite image according to a specified instruction. This subset contains 11K samples in total.
For Human-Centric, Object-Centric, and HOI tasks, the Parquet files share a unified schema with the following keys:
input_images: a list of input imagesoutput_image: the composed imageinstruction: a descriptive or imperative instruction specifying how to compose the input imagesseparate_prompt: captions corresponding to each input imageediting_type: a label describing the task typeFor the De&Re subset, the Parquet files contain the following keys:
reference: the real image used for decompositioninput: a list of images obtained by decomposing the reference imageoutput: the recomposed imageinstruction: a description of how to compose the decomposed input imagesseparate_prompt: captions corresponding to each decomposed input imageediting_type: a label describing the task type




If you find this dataset or the associated work useful for your research, please cite the paper:
@article{wei2025mico,
title={MICo-150K: A Comprehensive Dataset Advancing Multi-Image Composition},
author={Wei, Xinyu and Cen, Kangrui and Wei, Hongyang and Guo, Zhen and Li, Bairui and Wang, Zeqing and Zhang, Jinrui and Zhang, Lei},
journal={arXiv preprint arXiv:2512.07348},
year={2025}
}
51 commits
MICo-150K is a large-scale synthetic dataset generated by Nano Banana and Nano Banana Pro, designed to advance open-source models in Multi-Image Composition (MICo).
We fine-tune a diverse set of base models—including Qwen-Image, BAGEL, OmniGen2, Lumina-DiMOO, and BLIP3o-Next-Edit, on MICo-150K. All models demonstrate substantial performance improvements on our proposed MICo Bench after fine-tuning. Notably, Qwen-Image, originally developed as a text-to-image model, achieves remarkable gains after adaptation. The fine-tuned variant, Qwen-Image-MICo, surpasses Qwen-Image-Edit-2509 on both MICo Bench and OmniContext benchmarks, highlighting its strong generalization capability and broad applicability in multi-image composition scenarios.

We organize the MICo-150K dataset into three primary categories and one specialized task subset, De&Re. Each primary category encompasses multiple sub-tasks:
The De&Re subset focuses on decomposition and recomposition tasks. Specifically, elements from a real-world image (e.g., persons, apparel, scenes) are first decomposed into multiple component images. These components are subsequently recomposed into a single composite image according to a specified instruction. This subset contains 11K samples in total.
For Human-Centric, Object-Centric, and HOI tasks, the Parquet files share a unified schema with the following keys:
input_images: a list of input imagesoutput_image: the composed imageinstruction: a descriptive or imperative instruction specifying how to compose the input imagesseparate_prompt: captions corresponding to each input imageediting_type: a label describing the task typeFor the De&Re subset, the Parquet files contain the following keys:
reference: the real image used for decompositioninput: a list of images obtained by decomposing the reference imageoutput: the recomposed imageinstruction: a description of how to compose the decomposed input imagesseparate_prompt: captions corresponding to each decomposed input imageediting_type: a label describing the task type




If you find this dataset or the associated work useful for your research, please cite the paper:
@article{wei2025mico,
title={MICo-150K: A Comprehensive Dataset Advancing Multi-Image Composition},
author={Wei, Xinyu and Cen, Kangrui and Wei, Hongyang and Guo, Zhen and Li, Bairui and Wang, Zeqing and Zhang, Jinrui and Zhang, Lei},
journal={arXiv preprint arXiv:2512.07348},
year={2025}
}
51 commits