GAN-Based Data Augmentation Strategies for Imbalanced Industrial Datasets: A Comparative Study with Transfer Learning

Nour Dbouk Majd Saied Clovis Francis

Jul 2026 · Journal of Nondestructive Evaluation Diagnostics and Prognostics of Engineering Systems · Vol 9 · 0 citations

TL;DR

This work contributes by providing a benchmark for evaluating GAN-based augmentation methods in industrial inspection and by proposing a hybrid GAN–ResNet50 pipeline, which achieves superior classification accuracy through the integration of generative modeling with transfer learning.

Abstract

Automated surface defect detection in industrial manufacturing faces a critical challenge: the scarcity of balanced, annotated datasets severely limits the performance of deep learning models, particularly for minority defect classes that occur infrequently in production environments. While Generative Adversarial Networks (GANs) have emerged as promising tools for synthetic data generation, systematic comparative studies evaluating different GAN architectures and augmentation strategies for industrial defect detection remain limited. Furthermore, the integration of GAN-generated data with transfer learning techniques has not been thoroughly investigated. This study presents a comprehensive comparative analysis of four prominent GAN architectures, Deep Convolutional GAN (DCGAN), Conditional GAN (CGAN), Auxiliary Classifier GAN (ACGAN), and Wasserstein GAN with Gradient Penalty (WGAN-GP), applied to surface defect classification on a merged industrial dataset combining NEU-CLS and X-SSD (3,160 images across 13 defect classes). We systematically evaluate four augmentation strategies: uniform generation, selective minority oversampling, balanced augmentation, and distribution-aware generation. The generated synthetic images are integrated with transfer learning using a ResNet-50 backbone pretrained on ImageNet to assess downstream classification performance. Results show that GAN-based augmentation improves classification accuracy, particularly for minority defect types, however results depend strongly on the augmentation strategy. CGAN offered the best balance between accuracy, stability, and practical reliability. This work contributes by providing a benchmark for evaluating GAN-based augmentation methods in industrial inspection and by proposing a hybrid GAN–ResNet50 pipeline, which achieves superior classification accuracy through the integration of generative modeling with transfer learning.

DOI Semantic Scholar

View source

Similar papers

Open access 2026

Data-Efficient Semiconductor Wafer Defect Identification Using WGAN-GP Augmentation and Lightweight Deep Convolutional Networks

A data-efficient wafer defect classification framework that integrates generative data augmentation with lightweight deep learning models that enables reliable wafer defect classification in data-constrained manufacturing environments, making it well suited for deployment in resource-limited semiconductor fabrication f...

Adarsh Prakash, Umang Patwari, Swaroop Warade et al. · IEEE Access · 0 citations

Jul 2026

GAN-powered oversampling with strengthened sample distribution for class overlapping imbalanced data

This study introduces a novel approach, named SDCGAN, which oversamples with a Conditional Generative Adversarial Network (CGAN) where the generator is fed with strengthened distribution information extracted from a curated set of minority samples.

Yu-Lin Zhang, Xiao-Zhe Wang, Kai-Wen Xue et al. · International Journal of Dat... · 0 citations

Review Aug 2026

A Comprehensive Review of Generative Adversarial Networks in Deep Learning: Emerging Applications and Hybrid Architectures

Various strategies and improvements to enhance GANs stability and performance are examined, including hybrid architectures that integrate GANs with other deep learning models and practical utility in domain‐specific expert systems.

Kashif Iqbal, Xue Yu, Atifa Rafique et al. · Expert systems · 0 citations

Aug 2026

SMOG: an adaptive hybrid oversampling framework using SMOTE and conditional GAN for difficulty-aware learning on imbalanced data

SMOG, an adaptive hybrid oversampling framework that integrates the Synthetic Minority Over-sampling Technique with a Conditional Generative Adversarial Network (GAN)-based difficulty-aware learning strategy, highlights the effectiveness of adaptive hybrid generative strategies for intelligent learning on imbalanced da...

Jatinder Kaur, Vimal Parmar, B. K. Rao et al. · Evolutionary Intelligence · 0 citations

Open access Aug 2026

Ensemble attention deep network for imbalanced silicon wafer defect inspection

The proposed ensemble attention convolutional neural network (EACNN) first performs binary defect detection to distinguish between defective and non-defective samples, followed by a dedicated defect-type classification for fine-grained categorization that underscores EACNN’s ability to learn highly discriminative featu...

Uzma Batool, Zahid Iqbal, M. I. Shapiai et al. · PeerJ Computer Science · 1 citation

Open access 2026

FS-MSDANet: A Multi-Scale Dynamic Attention Network for Surface Defect Detection in Wear-Resistant Steel

A few-shot surface defect detection approach, termed Few-Shot Multi-Scale Dynamic Attention Network (FS-MSDANet), which performs well in accuracy and adeptly addresses challenges such as limited training samples, overfitting, and inadequate generalization is presented.

Jie Song, Gongxiang Cui, Chaoxiang Yang et al. · IEEE Access · 0 citations