Abstract:Recent breakthroughs in Neural Radiance Fields (NeRFs) have sparked significant demand for their integration into real-world 3D applications. However, the varied functionalities required by different 3D applications often necessitate diverse NeRF models with various pipelines, leading to tedious NeRF training for each target task and cumbersome trial-and-error experiments. Drawing inspiration from the generalization capability and adaptability of emerging foundation models, our work aims to develop one general-purpose NeRF for handling diverse 3D tasks. We achieve this by proposing a framework called Omni-Recon, which is capable of (1) generalizable 3D reconstruction and zero-shot multitask scene understanding, and (2) adaptability to diverse downstream 3D applications such as real-time rendering and scene editing. Our key insight is that an image-based rendering pipeline, with accurate geometry and appearance estimation, can lift 2D image features into their 3D counterparts, thus extending widely explored 2D tasks to the 3D world in a generalizable manner. Specifically, our Omni-Recon features a general-purpose NeRF model using image-based rendering with two decoupled branches: one complex transformer-based branch that progressively fuses geometry and appearance features for accurate geometry estimation, and one lightweight branch for predicting blending weights of source views. This design achieves state-of-the-art (SOTA) generalizable 3D surface reconstruction quality with blending weights reusable across diverse tasks for zero-shot multitask scene understanding. In addition, it can enable real-time rendering after baking the complex geometry branch into meshes, swift adaptation to achieve SOTA generalizable 3D understanding performance, and seamless integration with 2D diffusion models for text-guided 3D editing.

SG-NeRF: Neural Surface Reconstruction with Scene Graph Optimization

Robust SG-NeRF: Robust Scene Graph Aided Neural Surface Reconstruction

NeRS: Neural Reflectance Surfaces for Sparse-view 3D Reconstruction in the Wild

Enhance-NeRF: Multiple Performance Evaluation for Neural Radiance Fields

SurfelNeRF: Neural Surfel Radiance Fields for Online Photorealistic Reconstruction of Indoor Scenes

SiNeRF: Sinusoidal Neural Radiance Fields for Joint Pose Estimation and Scene Reconstruction

SG-NeRF: Semantic-guided Point-based Neural Radiance Fields

Fast Radiance Field Reconstruction from Sparse Inputs

CT-NeRF: Incremental Optimizing Neural Radiance Field and Poses with Complex Trajectory

Occlusion-Free Scene Recovery Via Neural Radiance Fields

SCALAR-NeRF: SCAlable LARge-scale Neural Radiance Fields for Scene Reconstruction

Omni-Recon: Harnessing Image-based Rendering for General-Purpose Neural Radiance Fields

Improving Neural Radiance Fields with Depth-aware Optimization for Novel View Synthesis

SG-NeRF: Sparse-Input Generalized Neural Radiance Fields for Novel View Synthesis

OmniNeRF: Hybriding Omnidirectional Distance and Radiance fields for Neural Surface Reconstruction

SGCNeRF: Few-Shot Neural Rendering via Sparse Geometric Consistency Guidance

GNeRP: Gaussian-guided Neural Reconstruction of Reflective Objects with Noisy Polarization Priors

NeRF-HuGS: Improved Neural Radiance Fields in Non-static Scenes Using Heuristics-Guided Segmentation

SN 2 eRF: A Framework for Neural Radiance Fields given Sparse and Noisy Poses

Improving Neural Radiance Field using Near-Surface Sampling with Point Cloud Generation