AquaSAM: Underwater Image Foreground Segmentation

TL;DR

AquaSAM adapts the Segment Anything Model for underwater image segmentation, significantly improving performance across various underwater targets by fine-tuning on specialized datasets.

Contribution

It introduces a simple fine-tuning approach to extend SAM's capabilities specifically for underwater foreground segmentation, a novel application.

Findings

AquaSAM outperforms default SAM on underwater segmentation tasks.

Achieves 7.13% DSC and 8.27% mIoU improvements.

Effective on challenging tasks like coral reef segmentation.

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.