RobustSAM: Segment Anything Robustly on Degraded Images

TL;DR

RobustSAM enhances the Segment Anything Model's ability to accurately segment degraded images by introducing minimal additional parameters and a new dataset, significantly improving zero-shot and downstream task performance.

Contribution

We propose RobustSAM, a method that improves SAM's robustness to degraded images with minimal parameter increase and introduce the Robust-Seg dataset for training and evaluation.

Findings

RobustSAM outperforms original SAM on degraded images.

It improves zero-shot segmentation accuracy.

Enhances downstream tasks like dehazing and deblurring.

Abstract

Segment Anything Model (SAM) has emerged as a transformative approach in image segmentation, acclaimed for its robust zero-shot segmentation capabilities and flexible prompting system. Nonetheless, its performance is challenged by images with degraded quality. Addressing this limitation, we propose the Robust Segment Anything Model (RobustSAM), which enhances SAM's performance on low-quality images while preserving its promptability and zero-shot generalization. Our method leverages the pre-trained SAM model with only marginal parameter increments and computational requirements. The additional parameters of RobustSAM can be optimized within 30 hours on eight GPUs, demonstrating its feasibility and practicality for typical research laboratories. We also introduce the Robust-Seg dataset, a collection of 688K image-mask pairs with different degradations designed to train and evaluate our model optimally. Extensive experiments across various segmentation tasks and datasets confirm RobustSAM's superior performance, especially under zero-shot conditions, underscoring its potential for extensive real-world application. Additionally, our method has been shown to effectively improve the performance of SAM-based downstream tasks such as single image dehazing and deblurring.