MureObjectStitch: Multi-reference Image Composition
This work addresses a specific challenge in image composition for computer vision applications, representing an incremental improvement.
The paper tackles the problem of preserving foreground details and adjusting pose/viewpoint in generative image composition by proposing a finetuning strategy with multi-reference images, achieving verified effectiveness on the MureCOM dataset.
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.