MureObjectStitch: Multi-reference Image Composition
Jiaxuan Chen, Bo Zhang, Qingdong He, Jinlong Peng, Li Niu

TL;DR
This paper introduces MureObjectStitch, a novel multi-reference image composition method that fine-tunes pretrained models with multiple images to improve foreground detail preservation and pose adjustment in composite images.
Contribution
It proposes a multi-reference finetuning strategy for generative image composition, enhancing realism and detail preservation in composite images.
Findings
Effective in preserving foreground details
Improves pose/viewpoint adjustment
Verified on MureCOM dataset
Abstract
Generative image composition aims to regenerate the given foreground object in the background image to produce a realistic composite image. The existing methods are struggling to preserve the foreground details and adjust the foreground pose/viewpoint at the same time. In this work, we propose an effective finetuning strategy for generative image composition model, in which we finetune a pretrained model using one or more images containing the same foreground object. Moreover, we propose a multi-reference strategy, which allows the model to take in multiple reference images of the foreground object. The experiments on MureCOM dataset verify the effectiveness of our method. The code and model have been released at https://github.com/bcmi/MureObjectStitch-Image-Composition.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsRadiomics and Machine Learning in Medical Imaging
