CERBERUS: Crack Evaluation & Recognition Benchmark for Engineering Reliability & Urban Stability

TL;DR

CERBERUS is a synthetic benchmark for training and evaluating AI models in detecting infrastructure cracks, combining realistic 3D scenarios and synthetic data to improve real-world defect detection performance.

Contribution

It introduces a versatile synthetic benchmark with 3D inspection scenarios and demonstrates the benefit of combining synthetic and real data for defect detection.

Findings

Combining synthetic and real data enhances detection accuracy.

The benchmark supports diverse inspection scenarios.

Synthetic data improves model robustness.

Abstract

CERBERUS is a synthetic benchmark designed to help train and evaluate AI models for detecting cracks and other defects in infrastructure. It includes a crack image generator and realistic 3D inspection scenarios built in Unity. The benchmark features two types of setups: a simple Fly-By wall inspection and a more complex Underpass scene with lighting and geometry challenges. We tested a popular object detection model (YOLO) using different combinations of synthetic and real crack data. Results show that combining synthetic and real data improves performance on real-world images. CERBERUS provides a flexible, repeatable way to test defect detection systems and supports future research in automated infrastructure inspection. CERBERUS is publicly available at https://github.com/justinreinman/Cerberus-Defect-Generator.