# From Restoration to Reconstruction: Rethinking 3D Gaussian Splatting for Underwater Scenes

arXiv: https://arxiv.org/abs/2509.17789 · HTML: https://mlsanity.com/paper/2509.17789/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 40 references that are in the map.
- First posted on arXiv 22 Sep 2025.

## Builds on (most cited first, 23 of 23)

- [Momentum Contrast for Unsupervised Visual Representation Learning](https://mlsanity.com/paper/1911.05722/) (2019)
- [3D Gaussian Splatting for Real-Time Radiance Field Rendering](https://mlsanity.com/paper/2308.04079/) (2023)
- [Instant Neural Graphics Primitives with a Multiresolution Hash Encoding](https://mlsanity.com/paper/2201.05989/) (2022)
- [Grounded SAM: Assembling Open-World Models for Diverse Visual Tasks](https://mlsanity.com/paper/2401.14159/) (2024)
- [Conditional Image Generation with PixelCNN Decoders](https://mlsanity.com/paper/1606.05328/) (2016)
- [WildGaussians: 3D Gaussian Splatting in the Wild](https://mlsanity.com/paper/2407.08447/) (2024)
- [Adding Gradient Noise Improves Learning for Very Deep Networks](https://mlsanity.com/paper/1511.06807/) (2015)
- [Underwater Image Enhancement via Medium Transmission-Guided Multi-Color Space Embedding](https://mlsanity.com/paper/2104.13015/) (2021)
- [Gaussian in the Wild: 3D Gaussian Splatting for Unconstrained Image Collections](https://mlsanity.com/paper/2403.15704/) (2024)
- [SeaThru-NeRF: Neural Radiance Fields in Scattering Media](https://mlsanity.com/paper/2304.07743/) (2023)
- [Contrastive Semi-supervised Learning for Underwater Image Restoration via Reliable Bank](https://mlsanity.com/paper/2303.09101/) (2023)
- [WaterSplatting: Fast Underwater 3D Scene Reconstruction Using Gaussian Splatting](https://mlsanity.com/paper/2408.08206/) (2024)
- [SeaSplat: Representing Underwater Scenes with 3D Gaussian Splatting and a Physically Grounded Image Formation Model](https://mlsanity.com/paper/2409.17345/) (2024)
- [Underwater Image Enhancement by Transformer-based Diffusion Model with Non-uniform Sampling for Skip Strategy](https://mlsanity.com/paper/2309.03445/) (2023)
- [WaterNeRF: Neural Radiance Fields for Underwater Scenes](https://mlsanity.com/paper/2209.13091/) (2022)
- [ScatterNeRF: Seeing Through Fog with Physically-Based Inverse Neural Rendering](https://mlsanity.com/paper/2305.02103/) (2023)
- [UW-GS: Distractor-Aware 3D Gaussian Splatting for Enhanced Underwater Scene Reconstruction](https://mlsanity.com/paper/2410.01517/) (2024)
- [Underwater Image Restoration via Contrastive Learning and a Real-world Dataset](https://mlsanity.com/paper/2106.10718/) (2021)
- [Aquatic-GS: A Hybrid 3D Representation for Underwater Scenes](https://mlsanity.com/paper/2411.00239/) (2024)
- [AquaNeRF: Neural Radiance Fields in Underwater Media with Distractor Removal](https://mlsanity.com/paper/2502.16351/) (2025)
- [Gaussian Splashing: Direct Volumetric Rendering Underwater](https://mlsanity.com/paper/2411.19588/) (2024)
- [Visual enhancement and 3D representation for underwater scenes: a review](https://mlsanity.com/paper/2505.01869/) (2025)
- [Bayesian Neural Networks for One-to-Many Mapping in Image Enhancement](https://mlsanity.com/paper/2501.14265/) (2025)

## Built on by (most cited first, 2 of 2)

- [OceanXL: Large-scale Underwater 3D Gaussian Splatting via Block Partitioning and Adaptive Pruning](https://mlsanity.com/paper/2609.29985/) (2026)
- [PUGS: Physics-Guided Uncertainty Modeling for Underwater Dynamic 3D Gaussian Splatting](https://mlsanity.com/paper/2610.17936/) (2026)

## Abstract (arXiv)

Underwater image degradation poses significant challenges for 3D reconstruction, where simplified physical models often fail in complex scenes. We propose R-Splatting, a unified framework that bridges underwater image restoration (UIR) with 3D Gaussian Splatting (3DGS) to improve both rendering quality and geometric fidelity. Our method integrates multiple enhanced views produced by diverse UIR models into a single reconstruction pipeline. During inference, a lightweight illumination generator samples latent codes to support diverse yet coherent renderings, while a contrastive loss ensures disentangled and stable illumination representations. Furthermore, we propose textitUncertainty-Aware Opacity Optimization (UAOO), which models opacity as a stochastic function to regularize training. This suppresses abrupt gradient responses triggered by illumination variation and mitigates overfitting to noisy or view-specific artifacts. Experiments on Seathru-NeRF and our new BlueCoral3D dataset demonstrate that R-Splatting outperforms strong baselines in both rendering quality and geometric accuracy.