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).

A knowledge distillation-based multi-scale relation-prototypical network for cross-domain few-shot defect classification

Knowledge Distillation Method for Surface Defect Detection.

A New Knowledge Distillation Network for Incremental Few-Shot Surface Defect Detection

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

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

DPNet: domain-aware prototypical network for interdisciplinary few-shot relation classification

Coarse-to-Fine Few-Shot Defect Recognition with Dynamic Weighting and Joint Metric

Hierarchical Multi-Scale Network for Cross-Scale Visual Defect Detection

Fabric defect classification using prototypical network of few-shot learning algorithm

TRNet: A Cross-Component Few-Shot Mechanical Fault Diagnosis

Self-supervised assisted multi-task learning network for one-shot defect segmentation with fake defect generation

Automated Detection of Defects with Low Semantic Information in X-ray Images Based on Deep Learning

ReNAP: Relation Network with Adaptive Prototypical Learning for Few-Shot Classification

Decision Fusion Network with Perception Fine-tuning for Defect Classification

Defect classification on limited labeled samples with multiscale feature fusion and semi-supervised learning

ODNet: A High Real-Time Network Using Orthogonal Decomposition for Few-Shot Strip Steel Surface Defect Classification

Prior knowledge-based residuals shrinkage prototype networks for cross-domain fault diagnosis

Zero-shot surface defect recognition with class knowledge graph

MFNet: A Novel Multilevel Feature Fusion Network With Multibranch Structure for Surface Defect Detection

Large Margin Prototypical Network for Few-shot Relation Classification with Fine-grained Features

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