Bibliographic Details
| Title: |
IMFine: 3D Inpainting via Geometry-guided Multi-view Refinement |
| Authors: |
Shi, Zhihao, Huo, Dong, Zhou, Yuhongze, Yin, Kejia, Min, Yan, Lu, Juwei, Zuo, Xinxin |
| Publication Year: |
2025 |
| Collection: |
Computer Science |
| Subject Terms: |
Computer Science - Computer Vision and Pattern Recognition |
| More Details: |
Current 3D inpainting and object removal methods are largely limited to front-facing scenes, facing substantial challenges when applied to diverse, "unconstrained" scenes where the camera orientation and trajectory are unrestricted. To bridge this gap, we introduce a novel approach that produces inpainted 3D scenes with consistent visual quality and coherent underlying geometry across both front-facing and unconstrained scenes. Specifically, we propose a robust 3D inpainting pipeline that incorporates geometric priors and a multi-view refinement network trained via test-time adaptation, building on a pre-trained image inpainting model. Additionally, we develop a novel inpainting mask detection technique to derive targeted inpainting masks from object masks, boosting the performance in handling unconstrained scenes. To validate the efficacy of our approach, we create a challenging and diverse benchmark that spans a wide range of scenes. Comprehensive experiments demonstrate that our proposed method substantially outperforms existing state-of-the-art approaches.
Comment: Accepted at CVPR 2025, \href{https://xinxinzuo2353.github.io/imfine/}{Project Page} |
| Document Type: |
Working Paper |
| Access URL: |
http://arxiv.org/abs/2503.04501 |
| Accession Number: |
edsarx.2503.04501 |
| Database: |
arXiv |
