Mirror-3DGS: Incorporating Mirror Reflections into 3D Gaussian Splatting
This addresses the challenge of modeling physical reflections in 3D scene reconstruction for applications like novel view synthesis, but it is incremental as it builds on existing 3DGS methods.
The paper tackled the problem of 3D Gaussian Splatting struggling with accurate mirror reflections, resulting in incorrect reconstructions, and introduced Mirror-3DGS to handle mirror geometries, which improved fidelity in novel view synthesis, surpassing Mirror-NeRF in mirror regions.
3D Gaussian Splatting (3DGS) has significantly advanced 3D scene reconstruction and novel view synthesis. However, like Neural Radiance Fields (NeRF), 3DGS struggles with accurately modeling physical reflections, particularly in mirrors, leading to incorrect reconstructions and inconsistent reflective properties. To address this challenge, we introduce Mirror-3DGS, a novel framework designed to accurately handle mirror geometries and reflections, thereby generating realistic mirror reflections. By incorporating mirror attributes into 3DGS and leveraging plane mirror imaging principles, Mirror-3DGS simulates a mirrored viewpoint from behind the mirror, enhancing the realism of scene renderings. Extensive evaluations on both synthetic and real-world scenes demonstrate that our method can render novel views with improved fidelity in real-time, surpassing the state-of-the-art Mirror-NeRF, especially in mirror regions.