PlaNeRF: SVD Unsupervised 3D Plane Regularization for NeRF Large-Scale Urban Scene Reconstruction

Neural Radiance Fields (NeRF) enable 3D scene reconstruction from 2D images and camera poses for Novel View Synthesis (NVS). Although NeRF can produce photorealistic results, it often suffers from overfitting to training views, leading to poor geometry reconstruction, especially in low-texture areas such as road surfaces in driving scenarios. This limitation restricts many important applications which require accurate geometry, such as extrapolated NVS, HD mapping, simulation and scene editing. To address this limitation, we propose a new method to improve NeRF's 3D structure using only RGB images and semantic maps. Our approach introduces a novel plane regularization based on Singular Value Decomposition (SVD), that does not rely on any geometric prior. In addition, we leverage the Structural Similarity Index Measure (SSIM) in patch-based loss design to properly initialize the volumetric representation of NeRF. Quantitative and qualitative results show that our method outperforms popular regularization approaches in accurate geometry reconstruction for large-scale outdoor scenes and achieves comparable rendering quality to SOTA methods on the KITTI-360 NVS benchmark.