AquaNeRF: Neural Radiance Fields in Underwater Media with Distractor Removal

2025-02-22

AI Summary

To address geometric distortion in underwater Neural Radiance Fields (NeRF) caused by dynamic scattering media—such as plankton and swimming organisms—this paper introduces the first single-ray, single-surface rendering framework tailored for underwater scenes. Methodologically: (1) a bias-augmented Gaussian weight function is designed to ensure consistent modeling of medium transmittance; (2) a depth-aware gradient scaling mechanism is introduced to enhance near-field geometric fidelity; and (3) an MLP-based implicit representation is integrated with a modified volumetric rendering formulation. Experiments demonstrate that the proposed method achieves +7.5% PSNR over Nerfacto and +6.2% over SeaThru-NeRF. It effectively suppresses underwater dynamic artifacts while preserving structural accuracy of static objects and fine background details. This work establishes a new paradigm for marine biological observation and high-fidelity underwater visual reconstruction.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

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.

Problem

• Underwater scene modeling
• Distractor removal
• Neural radiance field enhancement

Innovation

• NeRF renderer
• Gaussian weight function
• Depth-based scaling function