AquaNeRF: Neural Radiance Fields in Underwater Media with Distractor Removal
This work addresses the challenge of underwater scene reconstruction for applications like ocean life study and filming, representing an incremental improvement over existing NeRF methods.
The paper tackled the problem of modeling underwater scenes with Neural Radiance Fields (NeRF) by addressing visual artifacts like moving fish and suspended particles, resulting in a method that outperforms baselines by approximately 7.5% and 6.2% in PSNR and reduces artifacts while preserving details.
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.