Abstract:Visual surface defect inspection provides an important tool for product quality assessment in a wide range of industrial applications. In recent years, diverse convolutional neural network (CNN) models have been devel-oped for the high-accuracy image classification of natural objects. However, it remains a challenging task to develop deep-learning-based approaches for surface defect inspection of industrial products. The main challenge is that surface defects of industrial products occur in low probability, therefore it is impractical to build a large image dataset containing different types of defects to train/fine-tune deep CNN models to achieve satisfactory performance when applied to industrial inspection tasks. To overcome the above-mentioned problem, we build a light-weight defect classification model based on the pre-trained SqueezeNet architecture and present three effective techniques to achieve high-accuracy defect classification using very few collected defective training samples. First, we experimentally demonstrate that it is reasonable to freeze shallower convolutional layers and only fine-tune deeper layers in the last convolutional stage when there are not enough training samples. Second, we integrate batch-size independent Group Normalization (GN) into SqueezeNet to compute stable statistics based on a limited number of defective samples for training the classification model. Third, we add an auxiliary task to perform the similarity comparison of two defective samples during the training phase, significantly increasing the size of the training data and generating more distinctive defect-specific representations. The experimental results on the NEU-CLS dataset and USB-FS dataset verify the effectiveness of the proposed method. Our method can obtain accurate defect classification results when very few defective training samples are provided, i.e., about 97.69% and 82.92% average classification accuracy on the NEU-CLS and USB-FS datasets, respectively, under the 5-shot setting (5 images per defect category).

Segmentation and Detection of Industrial Surface Defect Based on Deep Neural Network

Small-scale defect detection in industrial environment based on lightweight deep learning network

AnomalySeg: Deep Learning-Based Fast Anomaly Segmentation Approach for Surface Defect Detection

Real-Time Defect Detection Model in Industrial Environment Based on Lightweight Deep Learning Network

An Effective Industrial Defect Classification Method under the Few-Shot Setting Via Two-Stream Training

Fully Convolutional Networks for Surface Defect Inspection in Industrial Environment

A Deep Learning-Based Surface Defect Inspection System Using Multiscale and Channel-Compressed Features

Surface Defect Detection of Industrial Parts Based on YOLOv5

Surface Defect Detection Algorithm Based on Feature-Enhanced YOLO

Surface defect detection of industrial components based on vision

Efficient surface defect detection using self-supervised learning strategy and segmentation network

An efficient lightweight convolutional neural network for industrial surface defect detection

Surface Defect Detection and Semantic Segmentation with a Novel Lightweight Deep Neural Network

Surface defect detection based on deep learning approach

A One‐Stage Deep Learning Model for Industrial Defect Detection

Detail-semantic guide network based on spatial attention for surface defect detection with fewer samples

A Two-Stage Industrial Defect Detection Framework Based on Improved-YOLOv5 and Optimized-Inception-ResnetV2 Models

Defect Detection of Mechanical Design Products Using Deep Neural Network

Research on deep learning method for rail surface defect detection

An end-to-end learning method for industrial defect detection

Deep Learning-Based Generic Automatic Surface Defect Inspection (ASDI) with Pixelwise Segmentation