EUVIMEDEUVIMED entdecken
Crossref

Inpaint-Anywhere: Zero-Shot Multi-Identity Inpainting with Efficient Diffusion Transformer

Junsheng Luan, Lei Zhao, Wei Xing

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (9) · 7644-7652

DOI: 10.1609/aaai.v40i9.37706 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Junsheng Luan und 2 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Subject-driven generation, which aims to synthesize visual content for a given identity V* with specific attributes, has garnered increasing attention in recent years. While existing methods demonstrate impressive identity consistency for both single and multiple identities, they often lack user-specified spatial control. Recent approaches, such as OminiControl-2 and EasyControl, enable inpainting conditioned on a single identity but fall short in multi-identity scenarios. In this paper, we introduce BoundID, a dataset synthesis pipeline for generating multi-identity images with bounding box annotations, and introduce Inpaint-Anywhere, a diffusion transformer framework for multi-identity inpainting. Given multiple identity references and corresponding masks, our method simultaneously generates all desired identities at precise locations while achieving both high identity and prompt fidelity. Extensive experiments show that Inpaint-Anywhere achieves state-of-the-art performance in multi-identity inpainting.

Zitieren

APA

Luan, J., Zhao, L., & Xing, W. (2026). Inpaint-Anywhere: Zero-Shot Multi-Identity Inpainting with Efficient Diffusion Transformer. Proceedings of the AAAI Conference on Artificial Intelligence, 40(9), 7644-7652. https://doi.org/10.1609/aaai.v40i9.37706

Vancouver

Luan J, Zhao L, Xing W. Inpaint-Anywhere: Zero-Shot Multi-Identity Inpainting with Efficient Diffusion Transformer. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(9):7644-7652. doi:10.1609/aaai.v40i9.37706

BibTeX
@article{Luan2026,
  author = {Luan, Junsheng and Zhao, Lei and Xing, Wei},
  title = {Inpaint-Anywhere: Zero-Shot Multi-Identity Inpainting with Efficient Diffusion Transformer},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {9},
  pages = {7644--7652},
  doi = {10.1609/aaai.v40i9.37706}
}
Kostenlos starten →