mlsanityData to 7 Oct 2026

3D-USE: From Image-Level to Scene-Level Underwater Enhancement

arXiv 2608.28020 ↗First posted 28 Aug 2026cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Underwater 3D reconstruction faithfully reproduces the color shifts and visibility loss of captured views, while physical inversion may leave estimation errors in the recovered scene appearance. We formulate Underwater Scene-level Enhancement (USE) as learning a persistent, visibility-enhanced 3D scene representation from degraded multi-view underwater observations, enabling consistent enhanced rendering. Realizing USE requires both a reliable scene representation for enhancement and a consistent enhancement target without paired enhanced 3D data. Therefore, we present 3D-USE, a two-stage framework. First, the Medium Radial Basis Anchor Representation (MediumRBF) establishes a medium-aware Gaussian scene by representing water effects with shared radial-basis anchors and explicitly decomposing object and medium contributions. Based on this fixed scene representation, Appearance Transition Consensus (ATC) transfers paired 2D underwater image enhancement (UIE) knowledge into scene-global and Gaussian-local targets, avoiding direct supervision from inconsistent enhanced views. An Underwater Bilateral Appearance Field (U-BAF) then realizes these targets in Gaussian radiance and medium appearance. The scene directly renders enhanced novel views without a 2D UIE model at inference. Experiments on real underwater scenes show improved visibility and cross-view consistency while preserving reconstruction quality.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 22 of its 35 references that are in the map. First posted on arXiv 28 Aug 2026.

Builds on

22 of 35 references.

References · 22

  1. 20233D Gaussian Splatting for Real-Time Radiance Field Renderingcited by 6,141
  2. 2020RAFT: Recurrent All-Pairs Field Transforms for Optical Flowcited by 2,385
  3. 2020NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collectionscited by 1,230
  4. 2021MUSIQ: Multi-scale Image Quality Transformercited by 1,058
  5. 2023Nerfstudio: A Modular Framework for Neural Radiance Field Developmentcited by 596
  6. 2024gsplat: An Open-Source Library for Gaussian Splattingcited by 279
  7. 2019An Underwater Image Enhancement Benchmark Dataset and Beyondcited by 250
  8. 2017Deep Bilateral Learning for Real-Time Image Enhancementcited by 200
  9. 2022StylizedNeRF: Consistent 3D Scene Stylization as Stylized NeRF via 2D-3D Mutual Learningcited by 140
  10. 2024Gaussian in the Wild: 3D Gaussian Splatting for Unconstrained Image Collectionscited by 117
  11. 2023SeaThru-NeRF: Neural Radiance Fields in Scattering Mediacited by 75
  12. 2022Underwater Ranker: Learn Which Is Better and How to Be Bettercited by 65
10 more
  1. 2024SWAG: Splatting in the Wild images with Appearance-conditioned Gaussianscited by 61
  2. 2023PUGAN: Physical Model-Guided Underwater Image Enhancement Using GAN with Dual-Discriminatorscited by 49
  3. 2024WaterSplatting: Fast Underwater 3D Scene Reconstruction Using Gaussian Splattingcited by 39
  4. 2024SeaSplat: Representing Underwater Scenes with 3D Gaussian Splatting and a Physically Grounded Image Formation Modelcited by 35
  5. 2024UW-GS: Distractor-Aware 3D Gaussian Splatting for Enhanced Underwater Scene Reconstructioncited by 23
  6. 2024UVEB: A Large-scale Benchmark and Baseline Towards Real-World Underwater Video Enhancementcited by 17
  7. 2025Plenodium: UnderWater 3D Scene Reconstruction with Plenoptic Medium Representationcited by 7
  8. 2022SUCRe: Leveraging Scene Structure for Underwater Color Restorationcited by 5
  9. 2026OceanSplat: Object-aware Gaussian Splatting with Trinocular View Consistency for Underwater Scene Reconstructioncited by 2
  10. 2026Spatiotemporal Degradation-Aware 3D Gaussian Splatting for Realistic Underwater Scene Reconstructioncited by 1

Built on by

Not cited yet.