AquaSAM: Underwater Image Foreground Segmentation

Muduo Xu, Jianhao Su, Yutao Liu
8/8/2023
cs.CV

No official code found. Best implementation: duooppa/AquaSAM (6 stars, MIT, last commit 2025-05-24).

Abstract

The Segment Anything Model (SAM) has revolutionized natural image segmentation, nevertheless, its performance on underwater images is still restricted. This work presents AquaSAM, the first attempt to extend the success of SAM on underwater images with the purpose of creating a versatile method for the segmentation of various underwater targets. To achieve this, we begin by classifying and extracting various labels automatically in SUIM dataset. Subsequently, we develop a straightforward fine-tuning method to adapt SAM to general foreground underwater image segmentation. Through extensive experiments involving eight segmentation tasks like human divers, we demonstrate that AquaSAM outperforms the default SAM model especially at hard tasks like coral reefs. AquaSAM achieves an average Dice Similarity Coefficient (DSC) of 7.13 (%) improvement and an average of 8.27 (%) on mIoU improvement in underwater segmentation tasks.

View on arXivView PDF

Code Implementations(1)

duooppa/AquaSAMCommunity100%
MIT

the first at-tempt to extend the success of SAM on underwater images with the purpose of creating a versatile tool for underwater image segmentation

61PythonJun 6, 20231 years agoMIT

Similar Papers

FSSUWNet: Mitigating the Fragility of Pre-trained Models with Feature Enhancement for Few-Shot Semantic Segmentation in Underwater Images

90% match1 repo32025

SparseUWSeg: Active Sparse Point-Label Augmentation for Underwater Semantic Segmentation

89% match2025

MV-Adapter: Enhancing Underwater Instance Segmentation via Adaptive Channel Attention

88% match2024

MARIS: Marine Open-Vocabulary Instance Segmentation with Geometric Enhancement and Semantic Alignment

88% match2025

Diving into Underwater: Segment Anything Model Guided Underwater Salient Instance Segmentation and A Large-scale Dataset

87% match1 repo1292024

Expose Camouflage in the Water: Underwater Camouflaged Instance Segmentation and Dataset

87% match1 repo132025

Cite this paper

@article{xu2023aquasam,
  title  = {AquaSAM: Underwater Image Foreground Segmentation},
  author = {Muduo Xu and Jianhao Su and Yutao Liu},
  year   = {2023},
  eprint = {2308.04218},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2308.04218v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2308.04218.svg)](https://codeofpaper.com/paper/2308.04218)

Discussion