BREPS: Bounding-Box Robustness Evaluation of Promptable Segmentation

TL;DR

This paper evaluates the robustness of promptable segmentation models like SAM to natural variations in bounding box prompts, introducing BREPS to generate adversarial prompts and benchmarking across diverse datasets.

Contribution

It presents BREPS, a novel white-box optimization method for adversarial bounding box generation, and provides a comprehensive robustness benchmark for promptable segmentation models.

Findings

SAM-like models are highly sensitive to natural prompt noise.

Adversarial bounding boxes can significantly alter segmentation quality.

Robustness varies across datasets and applications.

Abstract

Promptable segmentation models such as SAM have established a powerful paradigm, enabling strong generalization to unseen objects and domains with minimal user input, including points, bounding boxes, and text prompts. Among these, bounding boxes stand out as particularly effective, often outperforming points while significantly reducing annotation costs. However, current training and evaluation protocols typically rely on synthetic prompts generated through simple heuristics, offering limited insight into real-world robustness. In this paper, we investigate the robustness of promptable segmentation models to natural variations in bounding box prompts. First, we conduct a controlled user study and collect thousands of real bounding box annotations. Our analysis reveals substantial variability in segmentation quality across users for the same model and instance, indicating that SAM-like models are highly sensitive to natural prompt noise. Then, since exhaustive testing of all possible user inputs is computationally prohibitive, we reformulate robustness evaluation as a white-box optimization problem over the bounding box prompt space. We introduce BREPS, a method for generating adversarial bounding boxes that minimize or maximize segmentation error while adhering to naturalness constraints. Finally, we benchmark state-of-the-art models across 10 datasets, spanning everyday scenes to medical imaging. Code - https://github.com/emb-ai/BREPS.