Wat3R: Annotation-Free Underwater 3D Geometry Learning via Cross-Domain Semi-Supervised Adaptation

Forum topic · 小凯 · 2026-07-12

Summary

Wat3R is a cross-domain semi-supervised learning framework that adapts feed-forward 3D reconstruction models from aerial to underwater scenes without any annotated underwater data. Underwater 3D geometry estimation is hindered by light attenuation, scattering, and the lack of large-scale high-quality 3D annotations, making existing densely-supervised methods impractical. Wat3R uses a teacher-student architecture trained on abundant unlabeled real underwater video footage, and introduces a cross-view consistency loss that leverages geometric cues from other views to compensate for information degradation in the current view caused by water attenuation and scattering. The authors also build Water3D, an evaluation benchmark covering diverse water bodies and underwater scenes for geometry tasks. Experiments show Wat3R outperforms state-of-the-art methods in underwater multi-view depth estimation and point cloud reconstruction. Dataset and code are open-sourced at https://github.com/LSXI7/Wat3R (arXiv: 2607.08772).

Paper Overview

Research Area: Computer Vision (CV) Authors: Jiangwei Ren, Xingyu Jiang, Zijie Song, Wei Xu, Hongkai Lin, Dingkang Liang, Xiang Bai Published: 2026-07-09 arXiv: 2607.08772

Abstract

Estimating 3D geometry in underwater environments presents unique challenges due to light attenuation, scattering, and the absence of large-scale, high-quality 3D annotations. Pioneering methods rely on massive dense annotations that are impractical in underwater settings.

In this paper, the authors propose Wat3R, a cross-domain semi-supervised learning framework designed to adapt feed-forward 3D reconstruction models from air to underwater scenes.

Key Contributions

No annotated underwater data required: Wat3R follows a teacher-student architecture that learns robust geometry representations using only abundant unlabeled real underwater video footage.
Cross-view consistency loss: This loss leverages geometric cues from other views to compensate for information degradation in the current view caused by water attenuation and scattering.
Water3D dataset: Given the absence of comprehensive evaluation benchmarks, the authors construct Water3D, covering diverse water bodies and underwater scenes for evaluating geometry tasks.

Results

Experimental results show that Wat3R outperforms state-of-the-art methods in underwater multi-view depth estimation and point cloud reconstruction.

Resources

Code and dataset: https://github.com/LSXI7/Wat3R
arXiv: https://arxiv.org/abs/2607.08772

--- *Auto-collected on 2026-07-12.*