CXR-AD: Component X-ray Image Dataset for Industrial Anomaly Detection

2025-05-06

AI Summary

Existing anomaly detection benchmarks focus on visible-surface defects, lacking publicly available X-ray datasets for internal industrial component defects. Method: We introduce CXR-AD, the first open-source X-ray industrial anomaly detection dataset, comprising five real-world component types, 653 normal images, and 561 anomalous images with pixel-level defect masks. We systematically identify and formalize three core technical challenges in internal defect detection and propose a novel evaluation paradigm tailored to X-ray imaging characteristics. Leveraging real-world acquisition and meticulous annotation, we conduct benchmarking across three methodological categories: feature extraction, reconstruction-based modeling, and zero-shot learning. Results: State-of-the-art methods achieve 29.78% lower average performance on CXR-AD than on MVTec AD, confirming its heightened difficulty. CXR-AD establishes the first reproducible, scalable, industry-relevant benchmark for internal defect detection.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Machine Learning: Adversarial Learning & Robustness

Application Category

• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Internal defect detection constitutes a critical process in ensuring component quality, for which anomaly detection serves as an effective solution. However, existing anomaly detection datasets predominantly focus on surface defects in visible-light images, lacking publicly available X-ray datasets targeting internal defects in components. To address this gap, we construct the first publicly accessible component X-ray anomaly detection (CXR-AD) dataset, comprising real-world X-ray images. The dataset covers five industrial component categories, including 653 normal samples and 561 defect samples with precise pixel-level mask annotations. We systematically analyze the dataset characteristics and identify three major technical challenges: (1) strong coupling between complex internal structures and defect regions, (2) inherent low contrast and high noise interference in X-ray imaging, and (3) significant variations in defect scales and morphologies. To evaluate dataset complexity, we benchmark three state-of-the-art anomaly detection frameworks (feature-based, reconstruction-based, and zero-shot learning methods). Experimental results demonstrate a 29.78% average performance degradation on CXR-AD compared to MVTec AD, highlighting the limitations of current algorithms in handling internal defect detection tasks. To the best of our knowledge, CXR-AD represents the first publicly available X-ray dataset for component anomaly detection, providing a real-world industrial benchmark to advance algorithm development and enhance precision in internal defect inspection technologies.

Problem

• Lack of public X-ray datasets for internal defect detection
• Challenges in detecting defects in low-contrast noisy X-ray images
• Current algorithms underperform on complex internal defect tasks

Innovation

• First public X-ray dataset for internal defects
• Pixel-level mask annotations for defect samples
• Benchmarked three anomaly detection frameworks