SAFE-COLOR: Color Fidelity Benchmarks and Thresholds for Safety-Critical Object Detection

Color fidelity is often overlooked in simulation-based validation for autonomous vehicles, yet even minor color mismatches can undermine the reliability of AI-driven perception systems. In this paper, we systematically examine how controlled deviations in color reproduction-quantified by <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\Delta E$</tex> -affect object detection accuracy across 32 variants of YOLO. Using a Macbeth ColorChecker, we derive calibrations for key color transforms (brightness, contrast, hue, gamma, saturation and color bias) and apply these to the COCO validation set. Our evaluations demonstrate that increasing <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\Delta E$</tex> yields significant drops in detection metrics, especially for safety-critical categories such as pedestrians and cyclists. Based on these findings, we propose <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\Delta E$</tex> thresholds that define acceptable color fidelity in camera simulations (e.g., <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\Delta E\leq 3$</tex> for <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\Delta \mathbf{mAP}\leq 1\%)$</tex>. Further-more, we contribute these transformed datasets and scripts as a publicly available benchmark, enabling reproducible comparisons and guiding future research on color-based vulnerabilities in automated driving and other safety-critical domains.