TigAug: Data Augmentation for Testing Traffic Light Detection in Autonomous Driving Systems

2025-07-08

AI Summary

To address the high cost and limited diversity of manually collected data in testing autonomous driving traffic light detection models, this paper proposes a test-oriented meta-mutation-driven image augmentation method. We systematically formulate three interpretable categories of meta-mutations—weather conditions, camera imaging characteristics, and traffic light attributes—and design a family of naturalistic, targeted image transformations grounded in these relationships. Given only a small set of annotated images, our method automatically generates diverse synthetic samples for defect discovery and model retraining. Experiments across four state-of-the-art detectors and two benchmark datasets demonstrate that the proposed approach significantly improves fault detection rate (+23.6%), boosts mean Average Precision (mAP) of enhanced models by 2.1–4.8 points, and produces synthetically augmented images whose visual fidelity meets human acceptability standards.

Technology Category

• Computer Vision: Interpretability, Explainability, and Transparency
• Search and Optimization: Metareasoning and Metaheuristics
• Natural Language Processing: Safety and Robustness

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Autonomous vehicle technology has been developed in the last decades with recent advances in sensing and computing technology. There is an urgent need to ensure the reliability and robustness of autonomous driving systems (ADSs). Despite the recent achievements in testing various ADS modules, little attention has been paid on the automated testing of traffic light detection models in ADSs. A common practice is to manually collect and label traffic light data. However, it is labor-intensive, and even impossible to collect diverse data under different driving environments. To address these problems, we propose and implement TigAug to automatically augment labeled traffic light images for testing traffic light detection models in ADSs. We construct two families of metamorphic relations and three families of transformations based on a systematic understanding of weather environments, camera properties, and traffic light properties. We use augmented images to detect erroneous behaviors of traffic light detection models by transformation-specific metamorphic relations, and to improve the performance of traffic light detection models by retraining. Large-scale experiments with four state-of-the-art traffic light detection models and two traffic light datasets have demonstrated that i) TigAug is effective in testing traffic light detection models, ii) TigAug is efficient in synthesizing traffic light images, and iii) TigAug generates traffic light images with acceptable naturalness.

Problem

• Automated testing of traffic light detection models in autonomous driving systems
• Labor-intensive manual collection of diverse traffic light data
• Generating augmented traffic light images for testing and retraining models

Innovation

• Automated augmentation of labeled traffic light images
• Metamorphic relations and transformations for diverse conditions
• Improves model testing and performance via retraining