AquaNeRF: Neural Radiance Fields in Underwater Media with Distractor Removal

TL;DR

AquaNeRF introduces a novel underwater NeRF rendering method that effectively removes distractors like fish and particles, improving static scene reconstruction and reducing artifacts in underwater imaging.

Contribution

The paper presents a new NeRF renderer and optimization scheme tailored for underwater scenes, addressing unique artifacts and improving rendering quality over existing methods.

Findings

Outperforms baseline Nerfacto by 7.5% in PSNR

Outperforms SeaThru-NeRF by 6.2% in PSNR

Significantly reduces visual artifacts in underwater scene reconstructions

Abstract

Neural radiance field (NeRF) research has made significant progress in modeling static video content captured in the wild. However, current models and rendering processes rarely consider scenes captured underwater, which are useful for studying and filming ocean life. They fail to address visual artifacts unique to underwater scenes, such as moving fish and suspended particles. This paper introduces a novel NeRF renderer and optimization scheme for an implicit MLP-based NeRF model. Our renderer reduces the influence of floaters and moving objects that interfere with static objects of interest by estimating a single surface per ray. We use a Gaussian weight function with a small offset to ensure that the transmittance of the surrounding media remains constant. Additionally, we enhance our model with a depth-based scaling function to upscale gradients for near-camera volumes. Overall, our method outperforms the baseline Nerfacto by approximately 7.5\% and SeaThru-NeRF by 6.2% in terms of PSNR. Subjective evaluation also shows a significant reduction of artifacts while preserving details of static targets and background compared to the state of the arts.