ASSR-NeRF: Arbitrary-Scale Super-Resolution on Voxel Grid for High-Quality Radiance Fields Reconstruction

TL;DR

ASSR-NeRF introduces an attention-based 3D super-resolution framework that enhances high-quality radiance field reconstruction and novel view synthesis at arbitrary scales, addressing oversmoothing and multi-view consistency issues.

Contribution

The paper presents a novel VoxelGridSR model for direct 3D super-resolution on radiance fields, improving multi-view consistent high-resolution synthesis in unseen scenes.

Findings

Significant improvement in SRNVS quality demonstrated quantitatively.

Model generalizes well across diverse scenes.

Achieves multi-view consistent super-resolution.

Abstract

NeRF-based methods reconstruct 3D scenes by building a radiance field with implicit or explicit representations. While NeRF-based methods can perform novel view synthesis (NVS) at arbitrary scale, the performance in high-resolution novel view synthesis (HRNVS) with low-resolution (LR) optimization often results in oversmoothing. On the other hand, single-image super-resolution (SR) aims to enhance LR images to HR counterparts but lacks multi-view consistency. To address these challenges, we propose Arbitrary-Scale Super-Resolution NeRF (ASSR-NeRF), a novel framework for super-resolution novel view synthesis (SRNVS). We propose an attention-based VoxelGridSR model to directly perform 3D super-resolution (SR) on the optimized volume. Our model is trained on diverse scenes to ensure generalizability. For unseen scenes trained with LR views, we then can directly apply our VoxelGridSR to further refine the volume and achieve multi-view consistent SR. We demonstrate quantitative and qualitatively that the proposed method achieves significant performance in SRNVS.