Transfer learning for automated optical inspection

One of the challenges in applying convolutional neural networks to automated optical inspection is the lack of sufficient training data. In this paper we show that transfer learning can be successfully applied using image data from an entirely different domain. Focusing on optical inspection of texture images, we transfer weights from a source network trained with arbitrary unrelated images from the ImageNet dataset. Inspection experiments using our method show that one epoch of fine-tuning is sufficient to achieve 99.95% classification accuracy, while conventional transfer learning without fine-tuning achieves only 78.76%. An in-depth analysis of the effects of fine-tuning reveals that after fine-tuning, most of the unnecessary features encoded in the weights of the source network are deactivated, while meaningful features of the target data are amplified to capture new variations in the target domain.

[1]  Ivan Laptev,et al.  Learning and Transferring Mid-level Image Representations Using Convolutional Neural Networks , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[2]  Erhardt Barth,et al.  Non-parametric texture defect detection using Weibull features , 2011, Electronic Imaging.

[3]  Rob Fergus,et al.  Visualizing and Understanding Convolutional Networks , 2013, ECCV.

[4]  Bernd Scholz-Reiter,et al.  Automated surface inspection of cold-formed micro-parts , 2012 .

[5]  Trevor Darrell,et al.  DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition , 2013, ICML.

[6]  Paul J. Scott,et al.  Wavelets and their Applications for Surface Metrology , 2008 .

[7]  Ronald M. Summers,et al.  Deep Convolutional Neural Networks for Computer-Aided Detection: CNN Architectures, Dataset Characteristics and Transfer Learning , 2016, IEEE Transactions on Medical Imaging.

[8]  Hod Lipson,et al.  Understanding Neural Networks Through Deep Visualization , 2015, ArXiv.

[9]  Yoshua Bengio,et al.  How transferable are features in deep neural networks? , 2014, NIPS.

[10]  Reinhold Huber-Mörk,et al.  Convolutional Neural Networks for Steel Surface Defect Detection from Photometric Stereo Images , 2014, ISVC.

[11]  Franz Pernkopf,et al.  Visual Inspection of Machined Metallic High-Precision Surfaces , 2002, EURASIP J. Adv. Signal Process..

[12]  Jürgen Schmidhuber,et al.  Steel defect classification with Max-Pooling Convolutional Neural Networks , 2012, The 2012 International Joint Conference on Neural Networks (IJCNN).

[13]  Robert Sabourin,et al.  Transfer learning between texture classification tasks using Convolutional Neural Networks , 2015, 2015 International Joint Conference on Neural Networks (IJCNN).

[14]  Geoffrey E. Hinton,et al.  Visualizing Data using t-SNE , 2008 .

[15]  Bernd Scholz-Reiter,et al.  Design of deep convolutional neural network architectures for automated feature extraction in industrial inspection , 2016 .

[16]  Trevor Darrell,et al.  Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation , 2013, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[17]  P. S. Sastry,et al.  Bank of Weight Filters for Deep CNNs , 2016, ACML.

[18]  Pascal Vincent,et al.  Visualizing Higher-Layer Features of a Deep Network , 2009 .

[19]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[20]  Stefan Carlsson,et al.  CNN Features Off-the-Shelf: An Astounding Baseline for Recognition , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops.