SeaAnchor-GS: Identity-Anchored Gaussian Splatting for High-Fidelity Dynamic Underwater Scene Reconstruction
IEEE Signal Processing Letters · 2026 · 33 · 2450-2454
DOI: 10.1109/lsp.2026.3701832 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Signal Processing Letters, 2026. Verfasst von Yaoming Zhuang und 6 weiteren Autor:innen. Gefördert von: National Natural Science Foundation of China, Fundamental Research Funds for the Central Universities, Key R&D Program of National Fire Rescue Bureau.
Abstract (Originalsprache)
Dynamic underwater novel-view synthesis remains challenging because refraction, scattering, and particle interference undermine correspondence reliability, geometric consistency, and appearance stability. We propose SeaAnchor-GS, a robust dynamic 3D Gaussian splatting framework for underwater scene reconstruction. Each Gaussian is associated with a persistent identity embedding and transformed via identity–time conditioning, which improves deformation estimation under unstable underwater observations. A dual-branch residual dynamics module captures both dominant motion and fine-scale variations, while confidence-guided sampling, progressive deformation activation, and neighborhood-consistency regularization enhance optimization robustness and model compactness. Extensive experiments on dynamic underwater benchmarks show consistent improvements in reconstruction fidelity and perceptual quality, with a favorable balance between quality, efficiency, and representation compactness. Additional results on a static underwater benchmark suggest that the proposed representation remains competitive beyond the dynamic setting.
Förderung
- National Natural Science Foundation of China
- Fundamental Research Funds for the Central Universities
- Key R&D Program of National Fire Rescue Bureau
- Science and Technology Planning Project of Liaoning Province
Zitieren
APA
Zhuang, Y., Liu, T., Chao, Y., Li, L., Wu, H., Wu, C., & Liu, Z. (2026). SeaAnchor-GS: Identity-Anchored Gaussian Splatting for High-Fidelity Dynamic Underwater Scene Reconstruction. IEEE Signal Processing Letters, 33, 2450-2454. https://doi.org/10.1109/lsp.2026.3701832
Vancouver
Zhuang Y, Liu T, Chao Y, Li L, Wu H, Wu C, et al. SeaAnchor-GS: Identity-Anchored Gaussian Splatting for High-Fidelity Dynamic Underwater Scene Reconstruction. IEEE Signal Processing Letters. 2026;33:2450-2454. doi:10.1109/lsp.2026.3701832
BibTeX
@article{Zhuang2026,
author = {Zhuang, Yaoming and Liu, Tongrui and Chao, Yifan and Li, Li and Wu, Hao and Wu, Chengdong and Liu, Zhanlin},
title = {SeaAnchor-GS: Identity-Anchored Gaussian Splatting for High-Fidelity Dynamic Underwater Scene Reconstruction},
journal = {IEEE Signal Processing Letters},
year = {2026},
volume = {33},
pages = {2450--2454},
doi = {10.1109/lsp.2026.3701832}
}
EUVIMED