Abstract:Anomaly detection is a crucial means to detect unbalanced and multiclass abnormal data in industrial products. The existing anomaly detection model is usually suitable for products with a simple or regular background texture. However, there is a lack of a universal surface defect anomaly detection framework for railway products when facing the challenges: the complex background texture, huge intraclass variation, uneven distribution of interclass, and weak texture-featured defects. To fill this gap, an innovative generative adversarial network based on adaptive pyramid graph (APG) and variation residuals (APGVR-GAN) is proposed, aiming to improve the robustness of anomaly detection in railway products and other complex industrial supplies. First, the APG module is embedded in the encoder–decoder–encoder pattern, capturing the correlation description between neighbor regions, which is utilized to enhance the detection of abnormal defects with weak texture. Next, the variation residual module is employed to enhance the expression of various normal samples in the latent space and improve the identification ability for abnormal samples. Then, the dual-probability prototype loss is proposed to make different normal samples have more concentrated expression and more similar probability distribution centers in latent space. Finally, an adaptive focal-gate loss and a regularized log-likelihood loss are designed to overcome the imbalance problem in training samples with different background information. The effectiveness of the model is verified on three new railway datasets and three other industrial public datasets. The results show that in anomaly detection, APGVR-GAN on the three railway datasets, Northeastern University (NEU)-the rail surface defects dataset (RSDD)-1 dataset, NEU-RSDD-1 dataset, and RSDD area under curve (AUC) [that is the receiver operating characteristic curve (ROC)], can reach 0.89, 0.81, and 0.87, and its detection accuracy (ACC) reached 0.97, 0.96, and 0.92, which is much higher than other comparative anomaly detection algorithms. In the segmentation experiments on the RSDD dataset, compared with other known rail defect detection methods, the overall evaluation metric F-Score is 83.70 which is also better than other methods.

Generative Model-Driven Synthetic Training Image Generation: An Approach to Cognition in Railway Defect Detection

Generative Model-Driven Synthetic Training Image Generation: An Approach to Cognition in Rail Defect Detection

Railway Anomaly detection model using synthetic defect images generated by CycleGAN

Image Generation and Recognition for Railway Surface Defect Detection

Training a convolutional neural network for transportation sign detection using synthetic dataset

Controllable Image Synthesis of Industrial Data Using Stable Diffusion

An Adaptive Pyramid Graph and Variation Residual-Based Anomaly Detection Network for Rail Surface Defects

A Decoder-Free Reconstruction Method for Semi-Supervised Surface Defect Detection

ContRail: A Framework for Realistic Railway Image Synthesis using ControlNet

RailTrack-DaViT: A Vision Transformer-Based Approach for Automated Railway Track Defect Detection

Integrating Artificial Intelligence Models and Synthetic Image Data for Enhanced Asset Inspection and Defect Identification

Rail Surface Defect Detection Using A Transformer-Based Network

Intrusion Detection of Abnormal Objects for Railway Scenes Using Infrared Images

Rail surface defect data enhancement method based on improved ACGAN

DefectTwin: When LLM Meets Digital Twin for Railway Defect Inspection

Deep learning techniques to detect rail indications from ultrasonic data for automated rail monitoring and maintenance

Synthetic training data generation for deep learning based quality inspection

Data Augmentation of Railway Images for Track Inspection

Variational Autoencoder Diffusion Model (VAEDM) and Divergence Asynchronous Reinforcement Learning (DARL) for Rail Surface Defect Detection

Generative AI-based Pipeline Architecture for Increasing Training Efficiency in Intelligent Weed Control Systems

Advancing AI-Generated Image Detection: Enhanced Accuracy through CNN and Vision Transformer Models with Explainable AI Insights