IGFuse: Interactive 3D Gaussian Scene Reconstruction via Multi-Scans Fusion
This work solves the challenge of incomplete 3D scene reconstruction for computer vision and robotics applications, offering a scalable alternative to error-prone multi-stage pipelines.
The paper tackles the problem of reconstructing complete and interactive 3D scenes from multiple scans by addressing occlusions and limited sensor coverage, resulting in a framework that enables high-fidelity rendering and object-level manipulation without dense observations.
Reconstructing complete and interactive 3D scenes remains a fundamental challenge in computer vision and robotics, particularly due to persistent object occlusions and limited sensor coverage. Multiview observations from a single scene scan often fail to capture the full structural details. Existing approaches typically rely on multi stage pipelines, such as segmentation, background completion, and inpainting or require per-object dense scanning, both of which are error-prone, and not easily scalable. We propose IGFuse, a novel framework that reconstructs interactive Gaussian scene by fusing observations from multiple scans, where natural object rearrangement between captures reveal previously occluded regions. Our method constructs segmentation aware Gaussian fields and enforces bi-directional photometric and semantic consistency across scans. To handle spatial misalignments, we introduce a pseudo-intermediate scene state for unified alignment, alongside collaborative co-pruning strategies to refine geometry. IGFuse enables high fidelity rendering and object level scene manipulation without dense observations or complex pipelines. Extensive experiments validate the framework's strong generalization to novel scene configurations, demonstrating its effectiveness for real world 3D reconstruction and real-to-simulation transfer. Our project page is available online.