Improved Direct Voxel Grid Optimization for Radiance Fields Reconstruction

2022/06/10 by Sun, Cheng, Sun, Min, Chen, Hwann-Tzong · 4 citations
#FOS: Computer and information sciences #Graphics (cs.GR)

paper · doi:10.48550/arxiv.2206.05085

Abstract

In this technical report, we improve the DVGO framework (called DVGOv2), which is based on Pytorch and uses the simplest dense grid representation. First, we re-implement part of the Pytorch operations with cuda, achieving 2-3x speedup. The cuda extension is automatically compiled just in time. Second, we extend DVGO to support Forward-facing and Unbounded Inward-facing capturing. Third, we improve the space time complexity of the distortion loss proposed by mip-NeRF 360 from O(N²) to O(N). The distortion loss improves our quality and training speed. Our efficient implementation could allow more future works to benefit from the loss.

Cited by

• TextureSplat: Per-Primitive Texture Mapping for Reflective Gaussian Splatting
• Anti-Aliased 2D Gaussian Splatting
• FA-INR: Adaptive Implicit Neural Representations for Interpretable Exploration of Simulation Ensembles
• Hybrid 3D-4D Gaussian Splatting for Fast Dynamic Scene Representation

Related

• Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstruction
• VGOS: Voxel Grid Optimization for View Synthesis from Sparse Inputs
• Compressing Volumetric Radiance Fields to 1 MB
• Learning with Unreliability: Fast Few-shot Voxel Radiance Fields with Relative Geometric Consistency
• PyNeRF: Pyramidal Neural Radiance Fields