PDRF: Progressively Deblurring Radiance Field for Fast and Robust Scene Reconstruction from Blurry Images

TL;DR

PDRF introduces a progressive deblurring method for radiance fields that incorporates 3D scene context, enabling fast and robust scene reconstruction from blurry images, outperforming existing methods in speed and quality.

Contribution

The paper proposes a novel progressive deblurring scheme for radiance fields that models blur with 3D context and employs efficient importance sampling for rapid scene optimization.

Findings

PDRF is 15 times faster than previous state-of-the-art methods.

PDRF achieves better reconstruction quality on synthetic and real scenes.

The method effectively handles blurry images in scene reconstruction.

Abstract

We present Progressively Deblurring Radiance Field (PDRF), a novel approach to efficiently reconstruct high quality radiance fields from blurry images. While current State-of-The-Art (SoTA) scene reconstruction methods achieve photo-realistic rendering results from clean source views, their performances suffer when the source views are affected by blur, which is commonly observed for images in the wild. Previous deblurring methods either do not account for 3D geometry, or are computationally intense. To addresses these issues, PDRF, a progressively deblurring scheme in radiance field modeling, accurately models blur by incorporating 3D scene context. PDRF further uses an efficient importance sampling scheme, which results in fast scene optimization. Specifically, PDRF proposes a Coarse Ray Renderer to quickly estimate voxel density and feature; a Fine Voxel Renderer is then used to achieve high quality ray tracing. We perform extensive experiments and show that PDRF is 15X faster than previous SoTA while achieving better performance on both synthetic and real scenes.