CareCom: Generative Image Composition with Calibrated Reference Features
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (4) · 2877-2885
DOI: 10.1609/aaai.v40i4.37278 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Jiaxuan Chen und 4 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref).
Abstract (Originalsprache)
Image composition aims to seamlessly insert foreground object into background. Despite the huge progress in generative image composition, the existing methods are still struggling with simultaneous detail preservation and foreground pose/view adjustment. To address this issue, we extend the existing generative composition model to multi-reference version, which allows using arbitrary number of foreground reference images. Furthermore, we propose to calibrate the global and local features of foreground reference images to make them compatible with the background information. The calibrated reference features can supplement the original reference features with useful global and local information of proper pose/view. Extensive experiments on MVImgNet and MureCom demonstrate that the generative model can greatly benefit from the calibrated reference features.
Zitieren
APA
Chen, J., Zhang, B., He, Q., Peng, J., & Niu, L. (2026). CareCom: Generative Image Composition with Calibrated Reference Features. Proceedings of the AAAI Conference on Artificial Intelligence, 40(4), 2877-2885. https://doi.org/10.1609/aaai.v40i4.37278
Vancouver
Chen J, Zhang B, He Q, Peng J, Niu L. CareCom: Generative Image Composition with Calibrated Reference Features. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(4):2877-2885. doi:10.1609/aaai.v40i4.37278
BibTeX
@article{Chen2026,
author = {Chen, Jiaxuan and Zhang, Bo and He, Qingdong and Peng, Jinlong and Niu, Li},
title = {CareCom: Generative Image Composition with Calibrated Reference Features},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {4},
pages = {2877--2885},
doi = {10.1609/aaai.v40i4.37278}
}
EUVIMED