The document proposes a product defect detection system based on one-class classification that involves only normal samples during training. It consists of three parts: 1) A convolutional autoencoder is used to extract image features. 2) Principal component analysis performs dimensionality reduction on the extracted features. 3) The reduced features are used to train a support vector data description model for one-class classification to detect defects during testing. The approach is evaluated on carpet images and shows improved performance over state-of-the-art methods.
FABRIC DEFECT DETECTION BASED ON IMPROVED FASTER RCNNijaia
In the production process of fabric, defect detection plays an important role in the control of product
quality. Consider that traditional manual fabric defect detection method are time-consuming and
inaccuracy, utilizing computer vision technology to automatically detect fabric defects can better fulfill the
manufacture requirement. In this project, we improved Faster RCNN with convolutional block attention
module (CBAM) to detect fabric defects. Attention module is introduced from graph neural network, it can
infer the attention map from the intermediate feature map and multiply the attention map to adaptively
refine the feature. This method improve the performance of classification and detection without increase
the computation-consuming. The experiment results show that Faster RCNN with attention module can
efficient improve the classification accuracy.
NEURAL NETWORKS WITH DECISION TREES FOR DIAGNOSIS ISSUEScscpconf
This paper presents a new idea for fault detection and isolation (FDI) technique which is applied to industrial system. This technique is based on Neural Networks fault-free and Faulty
behaviours Models (NNFMs). NNFMs are used for residual generation, while decision tree architecture is used for residual evaluation. The decision tree is realized with data collected
from the NNFM’s outputs and is used to isolate detectable faults depending on computed threshold. Each part of the tree corresponds to specific residual. With the decision tree, it
becomes possible to take the appropriate decision regarding the actual process behaviour by evaluating few numbers of residuals. In comparison to usual systematic evaluation of all
residuals, the proposed technique requires less computational effort and can be used for on line diagnosis. An application example is presented to illustrate and confirm the effectiveness and the accuracy of the proposed approach.
NEURAL NETWORKS WITH DECISION TREES FOR DIAGNOSIS ISSUEScsitconf
This paper presents a new idea for fault detection and isolation (FDI) technique which is
applied to industrial system. This technique is based on Neural Networks fault-free and Faulty
behaviours Models (NNFMs). NNFMs are used for residual generation, while decision tree
architecture is used for residual evaluation. The decision tree is realized with data collected
from the NNFM’s outputs and is used to isolate detectable faults depending on computed
threshold. Each part of the tree corresponds to specific residual. With the decision tree, it
becomes possible to take the appropriate decision regarding the actual process behaviour by
evaluating few numbers of residuals. In comparison to usual systematic evaluation of all
residuals, the proposed technique requires less computational effort and can be used for on line
diagnosis. An application example is presented to illustrate and confirm the effectiveness and
the accuracy of the proposed approach.
FABRIC DEFECT DETECTION BASED ON IMPROVED FASTER RCNNijaia
In the production process of fabric, defect detection plays an important role in the control of product
quality. Consider that traditional manual fabric defect detection method are time-consuming and
inaccuracy, utilizing computer vision technology to automatically detect fabric defects can better fulfill the
manufacture requirement. In this project, we improved Faster RCNN with convolutional block attention
module (CBAM) to detect fabric defects. Attention module is introduced from graph neural network, it can
infer the attention map from the intermediate feature map and multiply the attention map to adaptively
refine the feature. This method improve the performance of classification and detection without increase
the computation-consuming. The experiment results show that Faster RCNN with attention module can
efficient improve the classification accuracy.
NEURAL NETWORKS WITH DECISION TREES FOR DIAGNOSIS ISSUEScscpconf
This paper presents a new idea for fault detection and isolation (FDI) technique which is applied to industrial system. This technique is based on Neural Networks fault-free and Faulty
behaviours Models (NNFMs). NNFMs are used for residual generation, while decision tree architecture is used for residual evaluation. The decision tree is realized with data collected
from the NNFM’s outputs and is used to isolate detectable faults depending on computed threshold. Each part of the tree corresponds to specific residual. With the decision tree, it
becomes possible to take the appropriate decision regarding the actual process behaviour by evaluating few numbers of residuals. In comparison to usual systematic evaluation of all
residuals, the proposed technique requires less computational effort and can be used for on line diagnosis. An application example is presented to illustrate and confirm the effectiveness and the accuracy of the proposed approach.
NEURAL NETWORKS WITH DECISION TREES FOR DIAGNOSIS ISSUEScsitconf
This paper presents a new idea for fault detection and isolation (FDI) technique which is
applied to industrial system. This technique is based on Neural Networks fault-free and Faulty
behaviours Models (NNFMs). NNFMs are used for residual generation, while decision tree
architecture is used for residual evaluation. The decision tree is realized with data collected
from the NNFM’s outputs and is used to isolate detectable faults depending on computed
threshold. Each part of the tree corresponds to specific residual. With the decision tree, it
becomes possible to take the appropriate decision regarding the actual process behaviour by
evaluating few numbers of residuals. In comparison to usual systematic evaluation of all
residuals, the proposed technique requires less computational effort and can be used for on line
diagnosis. An application example is presented to illustrate and confirm the effectiveness and
the accuracy of the proposed approach.
Synthesis and Refinement of Artificial HVAC Sensor Data Intended for Supervis...IES VE
IES' David McCabe presented at the 9th International Conference Improving Energy Efficiency in Commercial Buildings and Smart Communities (IEECB&SC’16) in Frankfurt on 16th March 2016.
This presentation was in support of a paper published by IES R&D in conjunction with the EINSTEIN project. The paper can be viewed here: http://www.iesve.com/corporate/media-center/white-papers/general/hvac-afdd-jun2016.pdf
Corrosion Detection Using A.I : A Comparison of Standard Computer Vision Tech...csandit
In this paper we present a comparison between stand
ard computer vision techniques and Deep
Learning approach for automatic metal corrosion (ru
st) detection. For the classic approach, a
classification based on the number of pixels contai
ning specific red components has been
utilized. The code written in Python used OpenCV li
braries to compute and categorize the
images. For the Deep Learning approach, we chose Ca
ffe, a powerful framework developed at
“Berkeley Vision and Learning Center” (BVLC). The
test has been performed by classifying
images and calculating the total accuracy for the t
wo different approaches.
CORROSION DETECTION USING A.I. : A COMPARISON OF STANDARD COMPUTER VISION TEC...cscpconf
In this paper we present a comparison between standard computer vision techniques and Deep
Learning approach for automatic metal corrosion (rust) detection. For the classic approach, a
classification based on the number of pixels containing specific red components has been
utilized. The code written in Python used OpenCV libraries to compute and categorize the
images. For the Deep Learning approach, we chose Caffe, a powerful framework developed at
“Berkeley Vision and Learning Center” (BVLC). The test has been performed by classifying
images and calculating the total accuracy for the two different approaches.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 65, NO. 3, M.docxsheronlewthwaite
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 65, NO. 3, MARCH 2018 2727
Electric Locomotive Bearing Fault Diagnosis
Using a Novel Convolutional
Deep Belief Network
Haidong Shao, Hongkai Jiang , Member, IEEE, Haizhou Zhang, and Tianchen Liang
Abstract—Bearing fault diagnosis is of significance to
enhance the reliability and security of electric locomotive.
In this paper, a novel convolutional deep belief network
(CDBN) is proposed for bearing fault diagnosis. First, an
auto-encoder is used to compress data and reduce the di-
mension. Second, a novel CDBN is constructed with Gaus-
sian visible units to learn the representative features. Third,
exponential moving average is employed to improve the
performance of the constructed deep model. The proposed
method is applied to analyze experimental signals collected
from electric locomotive bearings. The results show that
the proposed method is more effective than the traditional
methods and standard deep learning methods.
Index Terms—Convolutional deep belief network (CDBN),
electric locomotive bearing, exponential moving average
(EMA), fault diagnosis, feature learning.
NOMENCLATURE
ANFIS Adaptive neuro fuzzy inference system.
ANN Artificial neural network.
BPNN Back propagation neural network.
CDBN Convolutional deep belief network.
CNN Convolutional neural network.
CRBM Convolutional restricted Boltzmann machine.
DAE Deep auto-encoder.
DBN Deep belief network.
EMA Exponential moving average.
FD Frequency domain.
PCA Principal component analysis.
RBM Restricted Boltzmann machine.
SVM Support vector machine.
TD Time domain.
Manuscript received January 13, 2017; revised April 24, 2017 and
June 26, 2017; accepted August 5, 2017. Date of publication August
25, 2017; date of current version December 15, 2017. This work was
supported in part by the National Natural Science Foundation of China
under Grant 51475368, in part by the Shanghai Engineering Research
Center of Civil Aircraft Health Monitoring Foundation of China under
Grant GCZX-2015-02, and in part by the Innovation Foundation for Doc-
tor Dissertation of Northwestern Polytechnical University under Grant
CX201710. (Corresponding author: Hongkai Jiang.)
The authors are with the School of Aeronautics, Northwestern Poly-
technical University, Xi’an 710072, China (e-mail: [email protected]
edu.cn; [email protected]; [email protected];
[email protected]).
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TIE.2017.2745473
I. INTRODUCTION
E LECTRIC locomotive is playing a more and more impor-tant role in the modern transportation. The key parts of
electric locomotive usually get various faults due to the harsh
operating conditions, which may result in great catastrophes.
Bearing is one of the most widely used components in elec-
tric locomotive [1]; thus, automatic and accurate fault diagnosis
techniques are critically needed to ensure the sa ...
In this study, an end-to-end human iris recognition system is presented to automatically identify individuals for high level of security purposes. The deep learning technology based new 2D convolutional neural network (CNN) model is introduced for extracting the features and classifying the iris patterns. Firstly, the iris dataset is collected, preprocessed and augmented. The dataset are expanded and enhanced using data augmentation, histogram equalization (HE) and contrast-limited adaptive histogram equalization (CLAHE) techniques. Secondly, the features of the iris patterns were extracted and classified using CNN. The structure of CNN comprises of convolutional layers and ReLu layers for extracting the features, pooling layers for reducing the parameters, fully connected layer and Softmax layer for classifying the extracted features into N classes. For the training process and updating the weights, the backpropagation algorithm and adaptive moment estimation Adam optimizer are used. The experimental results carried out based on a graphics processing unit (GPU) and using Matlab. The overall training accuracy of the introduced system was 95.33% with a consumption time of 17.59 minutes for training set. While the testing accuracy 100% with a consumption time of 12 seconds. The introduced iris recognition system has been successfully applied.
A F AULT D IAGNOSIS M ETHOD BASED ON S EMI - S UPERVISED F UZZY C-M EANS...IJCI JOURNAL
Machine learning approaches are generally adopted i
n many fields including data mining, image
processing, intelligent fault diagnosis etc. As a c
lassic unsupervised learning technology, fuzzy C-me
ans
cluster analysis plays a vital role in machine lear
ning based intelligent fault diagnosis. With the ra
pid
development of science and technology, the monitori
ng signal data is numerous and keeps growing fast.
Only typical fault samples can be obtained and labe
led. Thus, how to apply semi-supervised learning
technology in fault diagnosis is significant for gu
aranteeing the equipment safety. According to this,
a novel
fault diagnosis method based on semi-supervised fuz
zy C-means(SFCM) cluster analysis is proposed.
Experimental results on Iris data set and the steel
plates faults data set show that this method is su
perior to
traditional fuzzy C-means clustering analysis
Black Box Model based Self Healing Solution for Stuck at Faults in Digital Ci...IJECEIAES
The paper proposes a design strategy to retain the true nature of the output in the event of occurrence of stuck at faults at the interconnect levels of digital circuits. The procedure endeavours to design a combinational architecture which includes attributes to identify stuck at faults present in the intermediate lines and involves a healing mechanism to redress the same. The simulated fault injection procedure introduces both single as well as multiple stuck-at faults at the interconnect levels of a two level combinational circuit in accordance with the directives of a control signal. The inherent heal facility attached to the formulation enables to reach out the fault free output even in the presence of faults. The Modelsim based simulation results obtained for the Circuit Under Test [CUT] implemented using a Read Only Memory [ROM], proclaim the ability of the system to survive itself from the influence of faults. The comparison made with the traditional Triple Modular Redundancy [TMR] exhibits the superiority of the scheme in terms of fault coverage and area overhead.
implementation of area efficient high speed eddr architectureKumar Goud
Abstract-This project presents an EDDR design, based on the residue-and-quotient (RQ) code, to embed into motion estimation (ME) for video coding testing applications. An error in processing elements (PEs), i.e. key components of a ME, can be detected and recovered effectively by using the EDDR design. The proposed EDDR design for ME testing can detect errors and recover data with an acceptable area overhead and timing penalty. The functional verification and synthesis can be done by Xilinx ISE. That is when compare to the existing design the implemented design area and timing will be reduced.
Index Terms—Area overhead, data recovery, error detection, reliability, residue-and-quotient (RQ) code, Xilinx ISE
Convolutional neural network with binary moth flame optimization for emotion ...IAESIJAI
Electroencephalograph (EEG) signals have the ability of real-time reflecting brain activities. Utilizing the EEG signal for analyzing human emotional states is a common study. The EEG signals of the emotions aren’t distinctive and it is different from one person to another as every one of them has different emotional responses to same stimuli. Which is why, the signals of the EEG are subject dependent and proven to be effective for the subject dependent detection of the Emotions. For the purpose of achieving enhanced accuracy and high true positive rate, the suggested system proposed a binary moth flame optimization (BMFO) algorithm for the process of feature selection and convolutional neural networks (CNNs) for classifications. In this proposal, optimum features are chosen with the use of accuracy as objective function. Ultimately, optimally chosen features are classified after that with the use of a CNN for the purpose of discriminating different emotion states.
A novel ensemble model for detecting fake newsIAESIJAI
Due the growing proliferation of fake news over the past couple of years, our objective in this paper is to propose an ensemble model for the automatic classification of article news as being either real or fake. For this purpose, we opt for a blending technique that combines three models, namely bidirectional long short-term memory (Bi-LSTM), stochastic gradient descent classifier and ridge classifier. The implementation of the proposed model (i.e. BI-LSR) on real world datasets, has shown outstanding results. In fact, it achieved an accuracy score of 99.16%. Accordingly, this ensemble learning has proven to do perform better than individual conventional machine learning and deep learning models as well as many ensemble learning approaches cited in the literature.
More Related Content
Similar to Product defect detection based on convolutional autoencoder and one-class classification
Synthesis and Refinement of Artificial HVAC Sensor Data Intended for Supervis...IES VE
IES' David McCabe presented at the 9th International Conference Improving Energy Efficiency in Commercial Buildings and Smart Communities (IEECB&SC’16) in Frankfurt on 16th March 2016.
This presentation was in support of a paper published by IES R&D in conjunction with the EINSTEIN project. The paper can be viewed here: http://www.iesve.com/corporate/media-center/white-papers/general/hvac-afdd-jun2016.pdf
Corrosion Detection Using A.I : A Comparison of Standard Computer Vision Tech...csandit
In this paper we present a comparison between stand
ard computer vision techniques and Deep
Learning approach for automatic metal corrosion (ru
st) detection. For the classic approach, a
classification based on the number of pixels contai
ning specific red components has been
utilized. The code written in Python used OpenCV li
braries to compute and categorize the
images. For the Deep Learning approach, we chose Ca
ffe, a powerful framework developed at
“Berkeley Vision and Learning Center” (BVLC). The
test has been performed by classifying
images and calculating the total accuracy for the t
wo different approaches.
CORROSION DETECTION USING A.I. : A COMPARISON OF STANDARD COMPUTER VISION TEC...cscpconf
In this paper we present a comparison between standard computer vision techniques and Deep
Learning approach for automatic metal corrosion (rust) detection. For the classic approach, a
classification based on the number of pixels containing specific red components has been
utilized. The code written in Python used OpenCV libraries to compute and categorize the
images. For the Deep Learning approach, we chose Caffe, a powerful framework developed at
“Berkeley Vision and Learning Center” (BVLC). The test has been performed by classifying
images and calculating the total accuracy for the two different approaches.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 65, NO. 3, M.docxsheronlewthwaite
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 65, NO. 3, MARCH 2018 2727
Electric Locomotive Bearing Fault Diagnosis
Using a Novel Convolutional
Deep Belief Network
Haidong Shao, Hongkai Jiang , Member, IEEE, Haizhou Zhang, and Tianchen Liang
Abstract—Bearing fault diagnosis is of significance to
enhance the reliability and security of electric locomotive.
In this paper, a novel convolutional deep belief network
(CDBN) is proposed for bearing fault diagnosis. First, an
auto-encoder is used to compress data and reduce the di-
mension. Second, a novel CDBN is constructed with Gaus-
sian visible units to learn the representative features. Third,
exponential moving average is employed to improve the
performance of the constructed deep model. The proposed
method is applied to analyze experimental signals collected
from electric locomotive bearings. The results show that
the proposed method is more effective than the traditional
methods and standard deep learning methods.
Index Terms—Convolutional deep belief network (CDBN),
electric locomotive bearing, exponential moving average
(EMA), fault diagnosis, feature learning.
NOMENCLATURE
ANFIS Adaptive neuro fuzzy inference system.
ANN Artificial neural network.
BPNN Back propagation neural network.
CDBN Convolutional deep belief network.
CNN Convolutional neural network.
CRBM Convolutional restricted Boltzmann machine.
DAE Deep auto-encoder.
DBN Deep belief network.
EMA Exponential moving average.
FD Frequency domain.
PCA Principal component analysis.
RBM Restricted Boltzmann machine.
SVM Support vector machine.
TD Time domain.
Manuscript received January 13, 2017; revised April 24, 2017 and
June 26, 2017; accepted August 5, 2017. Date of publication August
25, 2017; date of current version December 15, 2017. This work was
supported in part by the National Natural Science Foundation of China
under Grant 51475368, in part by the Shanghai Engineering Research
Center of Civil Aircraft Health Monitoring Foundation of China under
Grant GCZX-2015-02, and in part by the Innovation Foundation for Doc-
tor Dissertation of Northwestern Polytechnical University under Grant
CX201710. (Corresponding author: Hongkai Jiang.)
The authors are with the School of Aeronautics, Northwestern Poly-
technical University, Xi’an 710072, China (e-mail: [email protected]
edu.cn; [email protected]; [email protected];
[email protected]).
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TIE.2017.2745473
I. INTRODUCTION
E LECTRIC locomotive is playing a more and more impor-tant role in the modern transportation. The key parts of
electric locomotive usually get various faults due to the harsh
operating conditions, which may result in great catastrophes.
Bearing is one of the most widely used components in elec-
tric locomotive [1]; thus, automatic and accurate fault diagnosis
techniques are critically needed to ensure the sa ...
In this study, an end-to-end human iris recognition system is presented to automatically identify individuals for high level of security purposes. The deep learning technology based new 2D convolutional neural network (CNN) model is introduced for extracting the features and classifying the iris patterns. Firstly, the iris dataset is collected, preprocessed and augmented. The dataset are expanded and enhanced using data augmentation, histogram equalization (HE) and contrast-limited adaptive histogram equalization (CLAHE) techniques. Secondly, the features of the iris patterns were extracted and classified using CNN. The structure of CNN comprises of convolutional layers and ReLu layers for extracting the features, pooling layers for reducing the parameters, fully connected layer and Softmax layer for classifying the extracted features into N classes. For the training process and updating the weights, the backpropagation algorithm and adaptive moment estimation Adam optimizer are used. The experimental results carried out based on a graphics processing unit (GPU) and using Matlab. The overall training accuracy of the introduced system was 95.33% with a consumption time of 17.59 minutes for training set. While the testing accuracy 100% with a consumption time of 12 seconds. The introduced iris recognition system has been successfully applied.
A F AULT D IAGNOSIS M ETHOD BASED ON S EMI - S UPERVISED F UZZY C-M EANS...IJCI JOURNAL
Machine learning approaches are generally adopted i
n many fields including data mining, image
processing, intelligent fault diagnosis etc. As a c
lassic unsupervised learning technology, fuzzy C-me
ans
cluster analysis plays a vital role in machine lear
ning based intelligent fault diagnosis. With the ra
pid
development of science and technology, the monitori
ng signal data is numerous and keeps growing fast.
Only typical fault samples can be obtained and labe
led. Thus, how to apply semi-supervised learning
technology in fault diagnosis is significant for gu
aranteeing the equipment safety. According to this,
a novel
fault diagnosis method based on semi-supervised fuz
zy C-means(SFCM) cluster analysis is proposed.
Experimental results on Iris data set and the steel
plates faults data set show that this method is su
perior to
traditional fuzzy C-means clustering analysis
Black Box Model based Self Healing Solution for Stuck at Faults in Digital Ci...IJECEIAES
The paper proposes a design strategy to retain the true nature of the output in the event of occurrence of stuck at faults at the interconnect levels of digital circuits. The procedure endeavours to design a combinational architecture which includes attributes to identify stuck at faults present in the intermediate lines and involves a healing mechanism to redress the same. The simulated fault injection procedure introduces both single as well as multiple stuck-at faults at the interconnect levels of a two level combinational circuit in accordance with the directives of a control signal. The inherent heal facility attached to the formulation enables to reach out the fault free output even in the presence of faults. The Modelsim based simulation results obtained for the Circuit Under Test [CUT] implemented using a Read Only Memory [ROM], proclaim the ability of the system to survive itself from the influence of faults. The comparison made with the traditional Triple Modular Redundancy [TMR] exhibits the superiority of the scheme in terms of fault coverage and area overhead.
implementation of area efficient high speed eddr architectureKumar Goud
Abstract-This project presents an EDDR design, based on the residue-and-quotient (RQ) code, to embed into motion estimation (ME) for video coding testing applications. An error in processing elements (PEs), i.e. key components of a ME, can be detected and recovered effectively by using the EDDR design. The proposed EDDR design for ME testing can detect errors and recover data with an acceptable area overhead and timing penalty. The functional verification and synthesis can be done by Xilinx ISE. That is when compare to the existing design the implemented design area and timing will be reduced.
Index Terms—Area overhead, data recovery, error detection, reliability, residue-and-quotient (RQ) code, Xilinx ISE
Convolutional neural network with binary moth flame optimization for emotion ...IAESIJAI
Electroencephalograph (EEG) signals have the ability of real-time reflecting brain activities. Utilizing the EEG signal for analyzing human emotional states is a common study. The EEG signals of the emotions aren’t distinctive and it is different from one person to another as every one of them has different emotional responses to same stimuli. Which is why, the signals of the EEG are subject dependent and proven to be effective for the subject dependent detection of the Emotions. For the purpose of achieving enhanced accuracy and high true positive rate, the suggested system proposed a binary moth flame optimization (BMFO) algorithm for the process of feature selection and convolutional neural networks (CNNs) for classifications. In this proposal, optimum features are chosen with the use of accuracy as objective function. Ultimately, optimally chosen features are classified after that with the use of a CNN for the purpose of discriminating different emotion states.
A novel ensemble model for detecting fake newsIAESIJAI
Due the growing proliferation of fake news over the past couple of years, our objective in this paper is to propose an ensemble model for the automatic classification of article news as being either real or fake. For this purpose, we opt for a blending technique that combines three models, namely bidirectional long short-term memory (Bi-LSTM), stochastic gradient descent classifier and ridge classifier. The implementation of the proposed model (i.e. BI-LSR) on real world datasets, has shown outstanding results. In fact, it achieved an accuracy score of 99.16%. Accordingly, this ensemble learning has proven to do perform better than individual conventional machine learning and deep learning models as well as many ensemble learning approaches cited in the literature.
K-centroid convergence clustering identification in one-label per type for di...IAESIJAI
Disease prediction is a high demand field which requires significant support from machine learning (ML) to enhance the result efficiency. The research works on application of K-means clustering supervised classification in disease prediction where each class only has one labeled data. The K-centroid convergence clustering identification (KC3 I) system is based on semi-K-means clustering but only requires single labeled data per class for the training process with the training dataset to update the centroid. The KC3 I model also includes a dictionary box to index all the input centroids before and after the updating process. Each centroid matches with a corresponding label inside this box. After the training process, each time the input features arrive, the trained centroid will put them to its cluster depending on the Euclidean distance, then convert them into the specific class name, which is coherent to that centroid index. Two validation stages were carried out and accomplished the expectation in terms of precision, recall, F1-score, and absolute accuracy. The last part demonstrates the possibility of feature reduction by selecting the most crucial feature with the extra tree classifier method. Total data are fed into the KC3 I system with the most important features and remain the same accuracy.
Plant leaf detection through machine learning based image classification appr...IAESIJAI
Since maize is a staple diet for people, especially vegetarians and vegans, maize leaf disease has a significant influence here on the food industry including maize crop productivity. Therefore, it should be understood that maize quality must be optimal; yet, to do so, maize must be safeguarded from several illnesses. As a result, there is a great demand for such an automated system that can identify the condition early on and take the appropriate action. Early disease identification is crucial, but it also poses a major obstacle. As a result, in this research project, we adopt the fundamental k-nearest neighbor (KNN) model and concentrate on building and developing the enhanced k-nearest neighbor (EKNN) model. EKNN aids in identifying several classes of disease. To gather discriminative, boundary, pattern, and structurally linked information, additional high-quality fine and coarse features are generated. This information is then used in the classification process. The classification algorithm offers high-quality gradient-based features. Additionally, the proposed model is assessed using the Plant-Village dataset, and a comparison with many standard classification models using various metrics is also done.
Backbone search for object detection for applications in intrusion warning sy...IAESIJAI
In this work, we propose a novel backbone search method for object detection for applications in intrusion warning systems. The goal is to find a compact model for use in embedded thermal imaging cameras widely used in intrusion warning systems. The proposed method is based on faster region-based convolutional neural network (Faster R-CNN) because it can detect small objects. Inspired by EfficientNet, the sought-after backbone architecture is obtained by finding the most suitable width scale for the base backbone (ResNet50). The evaluation metrics are mean average precision (mAP), number of parameters, and number of multiply–accumulate operations (MACs). The experimental results showed that the proposed method is effective in building a lightweight neural network for the task of object detection. The obtained model can keep the predefined mAP while minimizing the number of parameters and computational resources. All experiments are executed elaborately on the person detection in intrusion warning systems (PDIWS) dataset.
Deep learning method for lung cancer identification and classificationIAESIJAI
Lung cancer (LC) is calming many lives and is becoming a serious cause of concern. The detection of LC at an early stage assists the chances of recovery. Accuracy of detection of LC at an early stage can be improved with the help of a convolutional neural network (CNN) based deep learning approach. In this paper, we present two methodologies for Lung cancer detection (LCD) applied on Lung image database consortium (LIDC) and image database resource initiative (IDRI) data sets. Classification of these LC images is carried out using support vector machine (SVM), and deep CNN. The CNN is trained with i) multiple batches and ii) single batch for LC image classification as non cancer and cancer image. All these methods are being implemented in MATLAB. The accuracy of classification obtained by SVM is 65%, whereas deep CNN produced detection accuracy of 80% and 100% respectively for multiple and single batch training. The novelty of our experimentation is near 100% classification accuracy obtained by our deep CNN model when tested on 25 Lung computed tomography (CT) test images each of size 512×512 pixels in less than 20 iterations as compared to the research work carried out by other researchers using cropped LC nodule images.
Optically processed Kannada script realization with Siamese neural network modelIAESIJAI
Optical character recognition (OCR) is a technology that allows computers to recognize and extract text from images or scanned documents. It is commonly used to convert printed or handwritten text into machine-readable format. This Study presents an OCR system on Kannada Characters based on siamese neural network (SNN). Here the SNN, a Deep neural network which comprises of two identical convolutional neural network (CNN) compare the script and ranks based on the dissimilarity. When lesser dissimilarity score is identified, prediction is done as character match. In this work the authors use 5 classes of Kannada characters which were initially preprocessed using grey scaling and convert it to pgm format. This is directly input into the Deep convolutional network which is learnt from matching and non-matching image between the CNN with contrastive loss function in Siamese architecture. The Proposed OCR system uses very less time and gives more accurate results as compared to the regular CNN. The model can become a powerful tool for identification, particularly in situations where there is a high degree of variation in writing styles or limited training data is available.
Embedded artificial intelligence system using deep learning and raspberrypi f...IAESIJAI
Melanoma is a kind of skin cancer that originates in melanocytes responsible for producing melanin, it can be a severe and potentially deadly form of cancer because it can metastasize to other regions of the body if not detected and treated early. To facilitate this process, Recently, various computer-assisted low-cost, reliable, and accurate diagnostic systems have been proposed based on artificial intelligence (AI) algorithms, particularly deep learning techniques. This work proposed an innovative and intelligent system that combines the internet of things (IoT) with a Raspberry Pi connected to a camera and a deep learning model based on the deep convolutional neural network (CNN) algorithm for real-time detection and classification of melanoma cancer lesions. The key stages of our model before serializing to the Raspberry Pi: Firstly, the preprocessing part contains data cleaning, data transformation (normalization), and data augmentation to reduce overfitting when training. Then, the deep CNN algorithm is used to extract the features part. Finally, the classification part with applied Sigmoid Activation Function. The experimental results indicate the efficiency of our proposed classification system as we achieved an accuracy rate of 92%, a precision of 91%, a sensitivity of 91%, and an area under the curve- receiver operating characteristics (AUC-ROC) of 0.9133.
Deep learning based biometric authentication using electrocardiogram and irisIAESIJAI
Authentication systems play an important role in wide range of applications. The traditional token certificate and password-based authentication systems are now replaced by biometric authentication systems. Generally, these authentication systems are based on the data obtained from face, iris, electrocardiogram (ECG), fingerprint and palm print. But these types of models are unimodal authentication, which suffer from accuracy and reliability issues. In this regard, multimodal biometric authentication systems have gained huge attention to develop the robust authentication systems. Moreover, the current development in deep learning schemes have proliferated to develop more robust architecture to overcome the issues of tradition machine learning based authentication systems. In this work, we have adopted ECG and iris data and trained the obtained features with the help of hybrid convolutional neural network- long short-term memory (CNN-LSTM) model. In ECG, R peak detection is considered as an important aspect for feature extraction and morphological features are extracted. Similarly, gabor-wavelet, gray level co-occurrence matrix (GLCM), gray level difference matrix (GLDM) and principal component analysis (PCA) based feature extraction methods are applied on iris data. The final feature vector is obtained from MIT-BIH and IIT Delhi Iris dataset which is trained and tested by using CNN-LSTM. The experimental analysis shows that the proposed approach achieves average accuracy, precision, and F1-core as 0.985, 0.962 and 0.975, respectively.
Hybrid channel and spatial attention-UNet for skin lesion segmentationIAESIJAI
Melanoma is a type of skin cancer which has affected many lives globally. The American Cancer Society research has suggested that it a serious type of skin cancer and lead to mortality but it is almost 100% curable if it is detected and treated in its early stages. Currently automated computer vision-based schemes are widely adopted but these systems suffer from poor segmentation accuracy. To overcome these issue, deep learning (DL) has become the promising solution which performs extensive training for pattern learning and provide better classification accuracy. However, skin lesion segmentation is affected due to skin hair, unclear boundaries, pigmentation, and mole. To overcome this issue, we adopt UNet based deep learning scheme and incorporated attention mechanism which considers low level statistics and high-level statistics combined with feedback and skip connection module. This helps to obtain the robust features without neglecting the channel information. Further, we use channel attention, spatial attention modulation to achieve the final segmentation. The proposed DL based scheme is instigated on publically available dataset and experimental investigation shows that the proposed Hybrid Attention UNet approach achieves average performance as 0.9715, 0.9962, 0.9710.
Photoplethysmogram signal reconstruction through integrated compression sensi...IAESIJAI
The transmission of photoplethysmogram (PPG) signals in real-time is extremely challenging and facilitates the use of an internet of things (IoT) environment for healthcare- monitoring. This paper proposes an approach for PPG signal reconstruction through integrated compression sensing and basis function aware shallow learning (CSBSL). Integrated-CSBSL approach for combined compression of PPG signals via multiple channels thereby improving the reconstruction accuracy for the PPG signals essential in healthcare monitoring. An optimal basis function aware shallow learning procedure is employed on PPG signals with prior initialization; this is further fine-tuned by utilizing the knowledge of various other channels, which exploit the further sparsity of the PPG signals. The proposed method for learning combined with PPG signals retains the knowledge of spatial and temporal correlation. The proposed Integrated-CSBSL approach consists of two steps, in the first step the shallow learning based on basis function is carried out through training the PPG signals. The proposed method is evaluated using multichannel PPG signal reconstruction, which potentially benefits clinical applications through PPG monitoring and diagnosis.
Speaker identification under noisy conditions using hybrid convolutional neur...IAESIJAI
Speaker identification is biometrics that classifies or identifies a person from other speakers based on speech characteristics. Recently, deep learning models outperformed conventional machine learning models in speaker identification. Spectrograms of the speech have been used as input in deep learning-based speaker identification using clean speech. However, the performance of speaker identification systems gets degraded under noisy conditions. Cochleograms have shown better results than spectrograms in deep learning-based speaker recognition under noisy and mismatched conditions. Moreover, hybrid convolutional neural network (CNN) and recurrent neural network (RNN) variants have shown better performance than CNN or RNN variants in recent studies. However, there is no attempt conducted to use a hybrid CNN and enhanced RNN variants in speaker identification using cochleogram input to enhance the performance under noisy and mismatched conditions. In this study, a speaker identification using hybrid CNN and the gated recurrent unit (GRU) is proposed for noisy conditions using cochleogram input. VoxCeleb1 audio dataset with real-world noises, white Gaussian noises (WGN) and without additive noises were employed for experiments. The experiment results and the comparison with existing works show that the proposed model performs better than other models in this study and existing works.
Multi-channel microseismic signals classification with convolutional neural n...IAESIJAI
Identifying and classifying microseismic signals is essential to warn of mines’ dangers. Deep learning has replaced traditional methods, but labor-intensive manual identification and varying deep learning outcomes pose challenges. This paper proposes a transfer learning-based convolutional neural network (CNN) method called microseismic signals-convolutional neural network (MS-CNN) to automatically recognize and classify microseismic events and blasts. The model was instructed on a limited sample of data to obtain an optimal weight model for microseismic waveform recognition and classification. A comparative analysis was performed with an existing CNN model and classical image classification models such as AlexNet, GoogLeNet, and ResNet50. The outcomes demonstrate that the MS-CNN model achieved the best recognition and classification effect (99.6% accuracy) in the shortest time (0.31 s to identify 277 images in the test set). Thus, the MS-CNN model can efficiently recognize and classify microseismic events and blasts in practical engineering applications, improving the recognition timeliness of microseismic signals and further enhancing the accuracy of event classification.
Sophisticated face mask dataset: a novel dataset for effective coronavirus di...IAESIJAI
Efficient and accurate coronavirus disease (COVID-19) surveillance necessitates robust identification of individuals wearing face masks. This research introduces the sophisticated face mask dataset (SFMD), a comprehensive compilation of high-quality face mask images enriched with detailed annotations on mask types, fits, and usage patterns. Leveraging cutting-edge deep learning models—EfficientNet-B2, ResNet50, and MobileNet-V2—, we compare SFMD against two established benchmarks: the real-world masked face dataset (RMFD) and the masked face recognition dataset (MFRD). Across all models, SFMD consistently outperforms RMFD and MFRD in key metrics, including accuracy, precision, recall, and F1 score. Additionally, our study demonstrates the dataset's capability to cultivate robust models resilient to intricate scenarios like low-light conditions and facial occlusions due to accessories or facial hair.
Transfer learning for epilepsy detection using spectrogram imagesIAESIJAI
Epilepsy stands out as one of the common neurological diseases. The neural activity of the brain is observed using electroencephalography (EEG). Manual inspection of EEG brain signals is a slow and arduous process, which puts heavy load on neurologists and affects their performance. The aim of this study is to find the best result of classification using the transfer learning model that automatically identify the epileptic and the normal activity, to classify EEG signals by using images of spectrogram which represents the percentage of energy for each coefficient of the continuous wavelet. Dataset includes the EEG signals recorded at monitoring unit of epilepsy used in this study to presents an application of transfer learning by comparing three models Alexnet, visual geometry group (VGG19) and residual neural network ResNet using different combinations with seven different classifiers. This study tested the models and reached a different value of accuracy and other metrics used to judge their performances, and as a result the best combination has been achieved with ResNet combined with support vector machine (SVM) classifier that classified EEG signals with a high success rate using multiple performance metrics such as 97.22% accuracy and 2.78% the value of the error rate.
Deep neural network for lateral control of self-driving cars in urban environ...IAESIJAI
The exponential growth of the automotive industry clearly indicates that self-driving cars are the future of transportation. However, their biggest challenge lies in lateral control, particularly in urban bottlenecking environments, where disturbances and obstacles are abundant. In these situations, the ego vehicle has to follow its own trajectory while rapidly correcting deviation errors without colliding with other nearby vehicles. Various research efforts have focused on developing lateral control approaches, but these methods remain limited in terms of response speed and control accuracy. This paper presents a control strategy using a deep neural network (DNN) controller to effectively keep the car on the centerline of its trajectory and adapt to disturbances arising from deviations or trajectory curvature. The controller focuses on minimizing deviation errors. The Matlab/Simulink software is used for designing and training the DNN. Finally, simulation results confirm that the suggested controller has several advantages in terms of precision, with lateral deviation remaining below 0.65 meters, and rapidity, with a response time of 0.7 seconds, compared to traditional controllers in solving lateral control.
Attention mechanism-based model for cardiomegaly recognition in chest X-Ray i...IAESIJAI
Recently, cardiovascular diseases (CVDs) have become a rapidly growing problem in the world, especially in developing countries. The latter are facing a lifestyle change that introduces new risk factors for heart disease, that requires a particular and urgent interest. Besides, cardiomegaly is a sign of cardiovascular diseases that refers to various conditions; it is associated with the heart enlargement that can be either transient or permanent depending on certain conditions. Furthermore, cardiomegaly is visible on any imaging test including Chest X-Radiation (X-Ray) images; which are one of the most common tools used by Cardiologists to detect and diagnose many diseases. In this paper, we propose an innovative deep learning (DL) model based on an attention module and MobileNet architecture to recognize Cardiomegaly patients using the popular Chest X-Ray8 dataset. Actually, the attention module captures the spatial relationship between the relevant regions in Chest X-Ray images. The experimental results show that the proposed model achieved interesting results with an accuracy rate of 81% which makes it suitable for detecting cardiomegaly disease.
Efficient commodity price forecasting using long short-term memory modelIAESIJAI
Predicting commodity prices, particularly food prices, is a significant concern for various stakeholders, especially in regions that are highly sensitive to commodity price volatility. Historically, many machine learning models like autoregressive integrated moving average (ARIMA) and support vector machine (SVM) have been suggested to overcome the forecasting task. These models struggle to capture the multifaceted and dynamic factors influencing these prices. Recently, deep learning approaches have demonstrated considerable promise in handling complex forecasting tasks. This paper presents a novel long short-term memory (LSTM) network-based model for commodity price forecasting. The model uses five essential commodities namely bread, meat, milk, oil, and petrol. The proposed model focuses on advanced feature engineering which involves moving averages, price volatility, and past prices. The results reveal that our model outperforms traditional methods as it achieves 0.14, 3.04%, and 98.2% for root mean square error (RMSE), mean absolute percentage error (MAPE), and R-squared (R2 ), respectively. In addition to the simplicity of the model, which consists of an LSTM single-cell architecture that reduced the training time to a few minutes instead of hours. This paper contributes to the economic literature on price prediction using advanced deep learning techniques as well as provides practical implications for managing commodity price instability globally.
1-dimensional convolutional neural networks for predicting sudden cardiacIAESIJAI
Sudden cardiac arrest (SCA) is a serious heart problem that occurs without symptoms or warning. SCA causes high mortality. Therefore, it is important to estimate the incidence of SCA. Current methods for predicting ventricular fibrillation (VF) episodes require monitoring patients over time, resulting in no complications. New technologies, especially machine learning, are gaining popularity due to the benefits they provide. However, most existing systems rely on manual processes, which can lead to inefficiencies in disseminating patient information. On the other hand, existing deep learning methods rely on large data sets that are not publicly available. In this study, we propose a deep learning method based on one-dimensional convolutional neural networks to learn to use discrete fourier transform (DFT) features in raw electrocardiogram (ECG) signals. The results showed that our method was able to accurately predict the onset of SCA with an accuracy of 96% approximately 90 minutes before it occurred. Predictions can save many lives. That is, optimized deep learning models can outperform manual models in analyzing long-term signals.
A deep learning-based approach for early detection of disease in sugarcane pl...IAESIJAI
In many regions of the nation, agriculture serves as the primary industry. The farming environment now faces a number of challenges to farmers. One of the major concerns, and the focus of this research, is disease prediction. A methodology is suggested to automate a process for identifying disease in plant growth and warning farmers in advance so they can take appropriate action. Disease in crop plants has an impact on agricultural production. In this work, a novel DenseNet-support vector machine: explainable artificial intelligence (DNet-SVM: XAI) interpretation that combines a DenseNet with support vector machine (SVM) and local interpretable model-agnostic explanation (LIME) interpretation has been proposed. DNet-SVM: XAI was created by a series of modifications to DenseNet201, including the addition of a support vector machine (SVM) classifier. Prior to using SVM to identify if an image is healthy or un-healthy, images are first feature extracted using a convolution network called DenseNet. In addition to offering a likely explanation for the prediction, the reasoning is carried out utilizing the visual cue produced by the LIME. In light of this, the proposed approach, when paired with its determined interpretability and precision, may successfully assist farmers in the detection of infected plants and recommendation of pesticide for the identified disease.
Generative AI Deep Dive: Advancing from Proof of Concept to ProductionAggregage
Join Maher Hanafi, VP of Engineering at Betterworks, in this new session where he'll share a practical framework to transform Gen AI prototypes into impactful products! He'll delve into the complexities of data collection and management, model selection and optimization, and ensuring security, scalability, and responsible use.
A tale of scale & speed: How the US Navy is enabling software delivery from l...sonjaschweigert1
Rapid and secure feature delivery is a goal across every application team and every branch of the DoD. The Navy’s DevSecOps platform, Party Barge, has achieved:
- Reduction in onboarding time from 5 weeks to 1 day
- Improved developer experience and productivity through actionable findings and reduction of false positives
- Maintenance of superior security standards and inherent policy enforcement with Authorization to Operate (ATO)
Development teams can ship efficiently and ensure applications are cyber ready for Navy Authorizing Officials (AOs). In this webinar, Sigma Defense and Anchore will give attendees a look behind the scenes and demo secure pipeline automation and security artifacts that speed up application ATO and time to production.
We will cover:
- How to remove silos in DevSecOps
- How to build efficient development pipeline roles and component templates
- How to deliver security artifacts that matter for ATO’s (SBOMs, vulnerability reports, and policy evidence)
- How to streamline operations with automated policy checks on container images
Elevating Tactical DDD Patterns Through Object CalisthenicsDorra BARTAGUIZ
After immersing yourself in the blue book and its red counterpart, attending DDD-focused conferences, and applying tactical patterns, you're left with a crucial question: How do I ensure my design is effective? Tactical patterns within Domain-Driven Design (DDD) serve as guiding principles for creating clear and manageable domain models. However, achieving success with these patterns requires additional guidance. Interestingly, we've observed that a set of constraints initially designed for training purposes remarkably aligns with effective pattern implementation, offering a more ‘mechanical’ approach. Let's explore together how Object Calisthenics can elevate the design of your tactical DDD patterns, offering concrete help for those venturing into DDD for the first time!
Smart TV Buyer Insights Survey 2024 by 91mobiles.pdf91mobiles
91mobiles recently conducted a Smart TV Buyer Insights Survey in which we asked over 3,000 respondents about the TV they own, aspects they look at on a new TV, and their TV buying preferences.
Le nuove frontiere dell'AI nell'RPA con UiPath Autopilot™UiPathCommunity
In questo evento online gratuito, organizzato dalla Community Italiana di UiPath, potrai esplorare le nuove funzionalità di Autopilot, il tool che integra l'Intelligenza Artificiale nei processi di sviluppo e utilizzo delle Automazioni.
📕 Vedremo insieme alcuni esempi dell'utilizzo di Autopilot in diversi tool della Suite UiPath:
Autopilot per Studio Web
Autopilot per Studio
Autopilot per Apps
Clipboard AI
GenAI applicata alla Document Understanding
👨🏫👨💻 Speakers:
Stefano Negro, UiPath MVPx3, RPA Tech Lead @ BSP Consultant
Flavio Martinelli, UiPath MVP 2023, Technical Account Manager @UiPath
Andrei Tasca, RPA Solutions Team Lead @NTT Data
GDG Cloud Southlake #33: Boule & Rebala: Effective AppSec in SDLC using Deplo...James Anderson
Effective Application Security in Software Delivery lifecycle using Deployment Firewall and DBOM
The modern software delivery process (or the CI/CD process) includes many tools, distributed teams, open-source code, and cloud platforms. Constant focus on speed to release software to market, along with the traditional slow and manual security checks has caused gaps in continuous security as an important piece in the software supply chain. Today organizations feel more susceptible to external and internal cyber threats due to the vast attack surface in their applications supply chain and the lack of end-to-end governance and risk management.
The software team must secure its software delivery process to avoid vulnerability and security breaches. This needs to be achieved with existing tool chains and without extensive rework of the delivery processes. This talk will present strategies and techniques for providing visibility into the true risk of the existing vulnerabilities, preventing the introduction of security issues in the software, resolving vulnerabilities in production environments quickly, and capturing the deployment bill of materials (DBOM).
Speakers:
Bob Boule
Robert Boule is a technology enthusiast with PASSION for technology and making things work along with a knack for helping others understand how things work. He comes with around 20 years of solution engineering experience in application security, software continuous delivery, and SaaS platforms. He is known for his dynamic presentations in CI/CD and application security integrated in software delivery lifecycle.
Gopinath Rebala
Gopinath Rebala is the CTO of OpsMx, where he has overall responsibility for the machine learning and data processing architectures for Secure Software Delivery. Gopi also has a strong connection with our customers, leading design and architecture for strategic implementations. Gopi is a frequent speaker and well-known leader in continuous delivery and integrating security into software delivery.
Builder.ai Founder Sachin Dev Duggal's Strategic Approach to Create an Innova...Ramesh Iyer
In today's fast-changing business world, Companies that adapt and embrace new ideas often need help to keep up with the competition. However, fostering a culture of innovation takes much work. It takes vision, leadership and willingness to take risks in the right proportion. Sachin Dev Duggal, co-founder of Builder.ai, has perfected the art of this balance, creating a company culture where creativity and growth are nurtured at each stage.
UiPath Test Automation using UiPath Test Suite series, part 4DianaGray10
Welcome to UiPath Test Automation using UiPath Test Suite series part 4. In this session, we will cover Test Manager overview along with SAP heatmap.
The UiPath Test Manager overview with SAP heatmap webinar offers a concise yet comprehensive exploration of the role of a Test Manager within SAP environments, coupled with the utilization of heatmaps for effective testing strategies.
Participants will gain insights into the responsibilities, challenges, and best practices associated with test management in SAP projects. Additionally, the webinar delves into the significance of heatmaps as a visual aid for identifying testing priorities, areas of risk, and resource allocation within SAP landscapes. Through this session, attendees can expect to enhance their understanding of test management principles while learning practical approaches to optimize testing processes in SAP environments using heatmap visualization techniques
What will you get from this session?
1. Insights into SAP testing best practices
2. Heatmap utilization for testing
3. Optimization of testing processes
4. Demo
Topics covered:
Execution from the test manager
Orchestrator execution result
Defect reporting
SAP heatmap example with demo
Speaker:
Deepak Rai, Automation Practice Lead, Boundaryless Group and UiPath MVP
UiPath Test Automation using UiPath Test Suite series, part 4
Product defect detection based on convolutional autoencoder and one-class classification
1. IAES International Journal of Artificial Intelligence (IJ-AI)
Vol. 12, No. 2, June 2023, pp. 912~920
ISSN: 2252-8938, DOI: 10.11591/ijai.v12.i2.pp912-920 912
Journal homepage: http://ijai.iaescore.com
Product defect detection based on convolutional autoencoder
and one-class classification
Meryem Chaabi1
, Mohamed Hamlich1
, Moncef Garouani1,2
1
Complex Cyber Physical System (CCPS) Laboratory, ENSAM, University of Hassan II, Casablanca, Morocco
2
University Littoral Cote d'Opale, UR 4491, LISIC, Laboratoire d'Informatique Signal et Image de la Cote d’Opale, Calais, France
Article Info ABSTRACT
Article history:
Received Sep 5, 2021
Revised Sep 16, 2022
Accepted Oct 16, 2022
To meet customer expectations and remain competitive, industrials try
constantly to improve their quality control systems. There is hence
increasing demand for adopting automatic defect detection solutions.
However, the biggest issue in addressing such systems is the imbalanced
aspect of industrial datasets. Often, defect-free samples far exceed the
defected ones, due to continuous improvement approaches adopted by
manufacturing companies. In this sense, we propose an automatic defect
detection system based on one-class classification (OCC) since it involves
only normal samples during training. It consists of three sub-models, first, a
convolutional autoencoder serves as latent features extractor, the extracted
features vectors are subsequently fed into the dimensionality reduction
process by performing principal component analysis (PCA), then the
reduced-dimensional data are used to train the one-class classifier support
vector data description (SVDD). During the test phase, both normal and
defected images are used. The first two stages of the trained model generate
a low-dimensional features vector, whereas the SVDD classifies the new
input, whether it is defect-free or defected. This approach is evaluated on the
carpet images from the industrial inspection dataset MVTec anomaly
detection (MVTec AD). During training, only normal images were used. The
results showed that the proposed method outperforms the state-of-the-art
methods.
Keywords:
Defect detection
Imbalanced data
MVTec anomaly detection
One-class classification
Support vector data description
This is an open access article under the CC BY-SA license.
Corresponding Author:
Meryem Chaabi
Complex Cyber Physical System (CCPS) Laboratory, ENSAM, University of Hassan II
Casablanca, Morocco
Email: meryem.chaabi-etu@etu.univh2c.ma
1. INTRODUCTION
Efficient quality inspection is one of the cornerstones of successful manufacturing companies. Since
human visual inspection is error-prone and subjective, there has been a movement towards automatic defect
detection systems [1], more than that, recently we witness the transition to the era of quality 4.0, which we
can define as a mature quality system that sought to leverage industry 4.0 technologies. In fact artificial
intelligence and machine learning [2] had proven their abilities to perform quality inspection [3]. A plethora
of defect detection methods have shown to be promising and efficient [4]. However most of them are based
on supervised learning [5], requiring considerable amount of normal and abnormal (i.e. defected) samples to
train the model efficiently, which is not always available in real applications. Actually, manufacturing
companies adopt increasingly improvement approaches [6] as lean management and six sigma [7], with the
attention of managing the fabrication process and reducing defective products by targeting zero defect
2. Int J Artif Intell ISSN: 2252-8938
Product defect detection based on convolutional autoencoder and one-class classification (Meryem Chaabi)
913
manufacturing. Defects hence rarely occur in production, and meanwhile there is enormous need to build
automatic defect detection system.
One-class classification (OCC) algorithms [8] offer the potential to create automatic inspection
system in early stage without having to wait for more defected samples to be collected. Consequently
industrials could exploit the important amount of data in their possession, despite highly skewed class
distribution. The main idea of OCC methods is to distinguish between normal and abnormal class, drew on
the knowledge gained of normal samples during training.
The term OCC was first introduced by [9] to denote a category of classification algorithms that
address cases where few to none defect samples are available for training; the normal class is well-defined
while abnormal one is under-sampled [10] which is quite common in industrial areas [11] ,and with that,
defects are seen as a deviation from defect-free class. The OCC concept encompasses several approaches,
such as methods based on density [12], distance [13], neural networks [14], [15], and boundary approaches
[16] that aims to encircle normal samples by a decision boundary. The work [17] developed adversarially
trained deep neural networks, the first component works as image reconstructor, while the second represents
the classifier. Autoencoders also were used to address OCC problems [18], [19]. In general, these researchers
assumed that autoencoders generates higher reconstruction error for defected samples. Going through
research works that address OCC, it is clearly apparent that support vector data description (SVDD) [20] is
one of the extensively used algorithms in OCC applications for its satisfactory results [21]. Luo et al. [22]
introduced a cost-sensitive SVDD, which is a method that provides different costs to classification errors.
The paper [23] developed a new SVDD approach to deal with uncertain data. They trained the SVDD by
using the similarity scores between examples and normal class. It is reported to outperforms regular SVDD in
terms of sensitivity to noise. Shi et al. [24] introduced improved SVDD by combining relative density weight
with SVDD. However, those SVDD based methods have difficulty dealing with large and high-dimensional
datasets [25] mainly because of optimization complexity.
2. METHOD
The intent of this work is to create an automatic defect detection system. We propose a model that
could manage imbalanced data and, at the same time yield to interesting results in high-dimensional spaces
using SVDD algorithm. In this section, we describe the proposed defect detection system in detail, and we
highlight the three components of our model: convolutional autoencoder, principal component analysis
algorithm (PCA) and SVDD algorithm.
2.1. Overview
Product surface images are used as input for our algorithm and only images of normal class are
present during training. The proposed system consists of three main phases. Firstly, we use a convolutional
autoencoder that allows extracting image’s abstract features. Once this submodel is trained, the decoder part
is discarded. Then the bottleneck features vector is later fed into PCA in order to perform dimensionality
reduction by inducing efficient and discriminating features representation. So that it serves as training input
of the SVDD classifier. The test images are forwarding through the trained model to determine whether the
image is normal or defected. The proposed system takes advantage of the convolutional autoencoder ability
to extract robust features automatically; meanwhile it alleviates the problem of SVDD of handling high-
dimensional datasets. Figure 1 shows the overview of the proposed approach, and Figure 2 illustrates the
flowchart of the entire system.
Figure 1. Shows the overview of the suggested approach
3. ISSN: 2252-8938
Int J Artif Intell, Vol. 12, No. 2, June 2023: 912-920
914
Figure 2. The flowchart of the proposed system
2.2. Features extraction
As a first stage a convolutional autoencoder (CAE) is used to extract image features. CAE can
automatically provide powerful feature learning [26]. Moreover, in [27] it has been proven that the more the
architecture of convolutional autoencoder becomes deeper, the more powerful features are extracted through
layers , in a way that first layer provides generic features while subsequent layers extract complex features.
Basically, autoencoder is a neural network that consists of two connected sub-models: an encoder and a
decoder. The symmetric structure of the autoencoder allows the reconstruction of input data through the
features vector provided by the bottleneck layer (i.e., encoder output) as illustrated in Figure 1. Considering
X the domain of the input data samples and Z the domain of the encodings.
Given a sample x ∈ 𝕏, z ∈ ℤ the encoded representation of x, and xr the reconstructed input. The
encoder-decoder process can be summarized,
𝑧 = 𝑔𝑒(𝑥; 𝜃𝑒) (1)
Feature extraction
Dimensionality reduction
Defect detection
Defect-free dataset for training
Set parameters initialization of CAE
Train the model
Compute gradients and update
backpropagation parameters
Compute Reconstruction error
Convergence
Determine eigenvalues and eigenvectors
Calculate the covariance matrix
Choose K eigenvectors corresponding to K
highest eigenvalues
Save the CAE parameters
Standardise data
Calculate the centre a and the radius R
Determine support vectors
Maximize the dual form of SVVD problem
Project input data into K-dimensional space
Define the parameter C and kernel width
Save SVDD parameters
Feature
Vectors
Reduced
dimensional
vector
Parameters of
SVDD
Test dataset
Trained encoder
Project feature vector into lower
dimensional space
Trained SVDD
Distance between
tested point and the
center ≤ R2
Defect-free
product
Defected
product
Epoch=epoch+1
Yes
No
4. Int J Artif Intell ISSN: 2252-8938
Product defect detection based on convolutional autoencoder and one-class classification (Meryem Chaabi)
915
𝑥𝑟 = 𝑔𝑑(z; 𝜃𝑑) (2)
Where ℊℯ: 𝕏 → ℤ is the encoding function and ℊ𝒹: ℤ → 𝕏 the decoding function, while 𝜃𝑒, 𝜃𝑑 represent the
encoder and decoder parameters respectively. A loss function is used to measure reconstruction error; we
have chosen ℓ2 loss for its simplicity and computational speed. The loss function is formulated as,
𝐿(x, 𝑥𝑟) = ‖x − 𝑥𝑟‖2
(3)
In this work, we trained the convolutional autoencoder on defect-free samples. Once the model is
trained, we use the bottleneck layer as automatic features extractor. The encoder consists of five convolution
layers where each layer is followed by a batch normalization layer. The max-pooling layer is used from the
second convolution layer where each layers use the rectifier linear unit (ReLu) acti-vation function,
i. The first two convolutional layers consist of 32 fil-ters of size 3×3, the second one followed by a
downsampling (max-pooling) layer.
ii. The third convolutional layer consists of 64 filters of size 3×3, followed by another downsampling
layer.
iii. The fourth convolutional layer consists of 128 fil-ters of size 3×3.
iv. The fifth convolutional layer consists of 256 filters of size 3×3.
2.3. Dimensionality reduction
Since SVDD has difficulty handling high-dimensional datasets, PCA [28] was applied to reduce the
dimensionality of features vectors provided by the CAE. PCA is a statistical procedure that aims to project
high-dimensional input data into a lower dimension space while retaining most of the information. Let Z be
an n×m data matrix, where the rows represent the n extracted vectors while the features are represented by
columns. PCA process is formulates,
a) Standardize the m-dimensional data
𝓏𝒾𝒿 ⇒
𝓏𝒾 𝒿 − 𝓊
𝓈𝒿
(4)
Here: 𝓏𝒾𝒿 is the (𝒾 𝒿)th entry of Z, 𝒾 = 1, … , n and 𝒿 = 1, … , m. 𝓊 and s𝒿 are respectively the mean
and the variance of 𝒿eme
dimension.
b) Calculate the covariance matrix Z𝓬
Z𝓬 =
1
𝑛
𝑍𝑇
𝑍 (5)
c) Construct the n eigenvalues and n eigenvectors via covariance matrix 𝑍𝒸
Z𝓬ν𝒾 = λ𝒾ν𝒾 (6)
Where 𝝀𝓲 denotes the eigenvalues, ν𝒾 represents the eigenvectors and 𝒾 = 1, … , n.
a. Choose k eigenvectors corresponding to the k highest eigenvalues k<m.
b. Project the input data into new k–dimensional space.
Let,
V = [ν1, ν2, … , νk] (7)
Then,
𝑧𝒾′ = 𝑉𝑇
𝑧 (8)
Where z𝒾′ is the low-dimensional features vector.
2.4. Defect detection
The defect detection task is done using SVDD. This algorithm attempts to determine a hypersphere
with the minimum volume that encircles almost all training data. The spherical boundary is characterized by
a center a and a radius R, hence during test, points that fall outside the boundary are considered as abnormal
as illustrated in Figure 3. The parameters R and a are defined by (9),
5. ISSN: 2252-8938
Int J Artif Intell, Vol. 12, No. 2, June 2023: 912-920
916
𝑚𝑖𝑛 R2
+ C ∑ ξ𝒾
𝓃
𝒾=1 (9)
Subject to,
‖z𝒾
′
− a‖ 2
≤ R2
+ ξ𝒾 ∀ 𝓲 (10)
ξ𝒾 ≥ 0 ∀ 𝓲 (11)
here 𝝃𝓲 : are slack variables that allow some points in training data to be outside the sphere and C represents a
penalty constant that controls the trade-off between the volume of the hypersphere and rejected points.
Figure 3. Illustration of SVDD
The optimization problem (9) is solved via its dual formulation,
max ∑ α𝔦〈z𝒾
′
, z𝒾
′〉 − ∑ α𝒾α𝒿〈z𝒾
′
, z𝒿
′〉
𝒾,𝒿
𝓃
𝒾=1 (12)
s.t.
0 ≤ α𝒾 ≤ C 𝒾=1, 2,…, 𝓃 (13)
∑ α𝒾 = 1
𝓷
𝓲=𝟏 (14)
Where α𝔦 are lagrange multipliers, and 〈z𝒾
′
, z𝒿
′ 〉 denotes the inner product of z𝒾
′
and z 𝒿
′
.The hypersphere
boundary is determined by support vectors for which 0 ≤ α𝒾 ≤ C. The center ‘a’ is calculated as,
a = ∑ z𝒾
′
𝓃
𝒾=1 α𝒾 (15)
The radius R is then computed by selecting an arbitrary support vector 𝓏𝓀
′
,
𝑅2
= ‖𝑧𝓀
′
− 𝒶‖2
= 〈𝑧𝓀
′
, 𝑧𝓀
′
〉 − 2 ∑ 𝛼𝔦〈𝑧𝓀
′
, 𝑧𝒾
′〉 + ∑ 𝛼𝒾𝛼𝒿〈𝑧𝒾
′
, 𝑧𝒿
′〉
𝒾,𝒿
𝓃
𝒾=1 (16)
For each new input 𝑦, the distance between this tested point and the center 𝒶 is computed as,
dist2(y) = 〈y , y〉 − 2 ∑ α𝔦〈y , z𝒾
′〉 + ∑ α𝒾α𝒿〈z𝒾
′
, z𝒿
′ 〉
𝒾,𝒿
𝓃
𝒾=1 (17)
Thus, points with dist2(y) ≥ R2
are considered as defected.
6. Int J Artif Intell ISSN: 2252-8938
Product defect detection based on convolutional autoencoder and one-class classification (Meryem Chaabi)
917
According to [20], replacing the inner product 〈z𝒾
′
, z𝒿
′ 〉 in (12) with appropriate kernel function
𝐾(𝓏𝒾
′
, 𝓏𝒿
′
)provides more flexibility to define the boundary. We have chosen to use the Gaussian kernel
function, given that it is the kernel reported to be yielding satisfactory results in OCC applications [29]. The
Gaussian kernel is defined as,
𝐾(z𝒾
′
, z𝒾
′
) = 𝑒𝑥𝑝
−‖z𝒾
′
−z𝒿
′
‖
2
2𝜎2 (18)
Where 𝜎 is the kernel width.
3. RESULTS AND DISCUSSION
The experimental environment is a computer with an i7-7700HQ CPU, 16GB of RAM, Nvidia RTX
2080 Ti GPU, running Windows 10, and we used the Tensorflow library to implement the proposed
approach. To demonstrate the effectiveness of our suggested system for defect detection, we conducted a set
of experiments on the carpet images from MVTec anomaly detection (MVTec AD) dataset [30],which is a
benchmark of natural images dedicated to unsupervised anomaly detection that mimics real industrial
inspection applications. The carpet dataset consists of 390 high-resolution images of 1024×1024 pixels,
divided into 72% of normal data and 28% of abnormal data .The training set is composed of defect-free
images, while the test set consists of normal images as well as images containing 5 different types of fine-
grained anomalies on the carpet’s surface like threads, cuts, metal contamination, color and holes. The
dataset overview is shown in Figure 4. Data augmentation was needed to diversify the training data set and
make the model more generalizable. To do so, random flips and rotations strategies were applied generating
1449 image. Finally, both categories were rescaled to 512×512 pixels.
Figure 4. Images of carpet dataset
We adopted the accuracy, area under the receiver operating characteristic curve (AUROC) and F1
score to measure the performance of the proposed method. Accuracy and F1 score are computed as,
𝑃𝑟𝑒𝑐𝑖𝑠𝑖𝑜𝑛 =
𝑇𝑃
𝑇𝑃+𝐹𝑃
(19)
𝑅𝑒𝑐𝑎𝑙𝑙 =
𝑇𝑃
𝑇𝑃+𝐹𝑁
(20)
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦 =
𝑇𝑃+𝑇𝑁
𝑇𝑃+𝑇𝑁+𝐹𝑃+𝐹𝑁
(21)
𝐹1𝑠𝑐𝑜𝑟𝑒 = 2.
𝑃𝑟𝑒𝑐𝑖𝑠𝑖𝑜𝑛.𝑅𝑒𝑐𝑎𝑙𝑙
𝑃𝑟𝑒𝑐𝑖𝑠𝑖𝑜𝑛+𝑅𝑒𝑐𝑎𝑙𝑙
(22)
Where true positive (TP): is the number of images correctly classified as defective.
True negative (TN): is the number of images correctly classified as normal.
7. ISSN: 2252-8938
Int J Artif Intell, Vol. 12, No. 2, June 2023: 912-920
918
False positive (FP): is the number of images incorrectly classified as defective.
False negative (FN): is the number of images incorrectly classified as normal.
For this implementation, we assumed that there are no outliers in training data. We defined C=1,
which implied that ξ_(i=)=0 [31]. The kernel width σ is optimized through cross validation. We investigated
the performance of our model on the carpet images and compare it with state-of-the-art methods.
Furthermore, we conducted comparison with the two-stage model “Convolutional autoencoder-SVDD”
(CAE-SVDD), to prove the importance of PCA algorithm in the proposed system.
As shown in Table 1, the proposed approach method outperforms the compared methods on all
metrics at detecting the defects in carpet dataset. The AUROC result of CAE-PCA-SVDD model was better,
with mean value 0.20% higher than the other algorithms. As can be observed in the comparison with
CAE-SVDD algorithm, applying PCA to reduce features vector dimensionality, leads to better performance
than the CAE-SVDD structure. Hence the proposed defect detection system could operate well in high
dimensional spaces. However, the proposed approach has some limitations. As shown in Figure 5, the system
failed to recognize a defect when the contrast of defected part is similar to the texture of the image.
Table 1. Experimental results of different approaches on the carpet dataset
Approach AUROC Accuracy F1 score
Liu et al. [18] 0.78 - -
Wang et al. [19] 0.94 0.71 0.88
Bergmann et al. [32] 0.87 0.67 0.87
Bergmann et al. [32] 0.59 0.50 0.54
CAE-SVDD 0.71 0.60 0.67
CAE-PCA-SVDD 0.98 0.94 0.97
Figure 5. A false negative case
4. CONCLUSION
The current paper proposes automatic defect detection system that attempts to mitigate the lack of
representative samples of defected images. Hence it can be applied in cases where significant amount of
normal data is available, while the defected class is characterized by few samples. The suggested approach
comprises three components: convolutional autoencoder for image features extraction, then PCA is applied to
reduce dimensionality of features vector, and finally SVDD is implemented to classify images. The
motivation behind this structure is to build automatic defect detection system that is able to handle effectively
imbalanced data. The proposed method has proved to be efficient in experiment developed on carpet dataset
from MVTec AD. It is important to note that SVDD parameters (C and the width 𝜎 ) vary from one dataset to
another. In this work C was defined as C=1, while 𝜎 was chosen through cross-validation to find best
accuracy. As a future work, we plan to evaluate the proposed approach on different datasets, and also utilize
an automatic method to determine the parameters: (C, 𝜎), which will facilitate the application of our
approach. Furthermore we will investigate further techniques to improve the ability of the proposed system to
handle texture features.
8. Int J Artif Intell ISSN: 2252-8938
Product defect detection based on convolutional autoencoder and one-class classification (Meryem Chaabi)
919
REFERENCES
[1] N. N. S. A. Rahman, N. M. Saad, A. R. Abdullah, M. R. M. Hassan, M. S. S. M. Basir, and N. S. M. Noor, “Automated real-time
vision quality inspection monitoring system,” Indonesian Journal of Electrical Engineering and Computer Science, vol. 11, no. 2,
pp. 775–783, 2018, doi: 10.11591/ijeecs.v11.i2.pp775-783.
[2] M. Hamlich and M. Ramdani, “Improved ant colony algorithms for data classification,” Proceedings of 2012 International
Conference on Complex Systems, ICCS 2012, 2012, doi: 10.1109/ICoCS.2012.6458563.
[3] A. Abdullahi, N. A. Samsudin, M. R. Ibrahim, M. S. Aripin, S. K. A. Khalid, and Z. A. Othman, “Towards IR4.0 implementation
in e-manufacturing: artificial intelligence application in steel plate fault detection,” Indonesian Journal of Electrical Engineering
and Computer Science, vol. 20, no. 1, pp. 430–436, 2020, doi: 10.11591/ijeecs.v20.i1.pp430-436.
[4] N. M. Saad, A. R. Abdullah, W. H. W. Hasan, N. N. S. A. Rahman, N. H. Ali, and I. N. Abdullah, “Automated vision based
defect detection using gray level co-occurrence matrix for beverage manufacturing industry,” IAES International Journal of
Artificial Intelligence, vol. 10, no. 4, pp. 818–829, 2021, doi: 10.11591/ijai.v10.i4.pp818-829.
[5] Z. Ren, F. Fang, N. Yan, and Y. Wu, “State of the art in defect detection based on machine vision,” International Journal of
Precision Engineering and Manufacturing - Green Technology, vol. 9, no. 2, pp. 661–691, 2022, doi: 10.1007/s40684-021-00343-6.
[6] G. Marodin, A. G. Frank, G. L. Tortorella, and T. Netland, “Lean product development and lean manufacturing: testing
moderation effects,” International Journal of Production Economics, vol. 203, pp. 301–310, 2018, doi: 10.1016/j.ijpe.2018.07.009.
[7] M. Smętkowska and B. Mrugalska, “Using six sigma DMAIC to improve the quality of the production process: a case study,”
Procedia - Social and Behavioral Sciences, vol. 238, pp. 590–596, 2018, doi: 10.1016/j.sbspro.2018.04.039.
[8] A. Mujeeb, W. Dai, M. Erdt, and A. Sourin, “One class based feature learning approach for defect detection using deep
autoencoders,” Advanced Engineering Informatics, vol. 42, 2019, doi: 10.1016/j.aei.2019.100933.
[9] L. D. Moya, MM and Koch, MW and Hostetler, “One-class classifier networks for target recognition applications,” NASA
STI/Recon Technical Report N, vol. 93, p. 24043, 1993.
[10] M. Chaabi and M. Hamlich, “A sight on defect detection methods for imbalanced industrial data,” ITM Web of Conferences,
vol. 43, p. 1012, 2022, doi: 10.1051/itmconf/20224301012.
[11] M. Garouani, A. Ahmad, M. Bouneffa, M. Hamlich, G. Bourguin, and A. Lewandowski, “Using meta-learning for automated
algorithms selection and configuration: an experimental framework for industrial big data,” Journal of Big Data, vol. 9, no. 1,
2022, doi: 10.1186/s40537-022-00612-4.
[12] J. Hamidzadeh and M. Moradi, “Improved one-class classification using filled function,” Applied Intelligence, vol. 48, no. 10,
pp. 3263–3279, 2018, doi: 10.1007/s10489-018-1145-y.
[13] J. Wang, W. Liu, K. Qiu, H. Xiong, and L. Zhao, “Dynamic hypersphere SVDD without describing boundary for one-class
classification,” Neural Computing and Applications, vol. 31, no. 8, pp. 3295–3305, 2019, doi: 10.1007/s00521-017-3277-0.
[14] X. E. Pantazi, D. Moshou, and A. A. Tamouridou, “Automated leaf disease detection in different crop species through image
features analysis and one class classifiers,” Computers and Electronics in Agriculture, vol. 156, pp. 96–104, 2019,
doi: 10.1016/j.compag.2018.11.005.
[15] M. Garouani, A. Ahmad, M. Bouneffa, M. Hamlich, G. Bourguin, and A. Lewandowski, “Towards big industrial data mining
through explainable automated machine learning,” International Journal of Advanced Manufacturing Technology, vol. 120,
no. 1–2, pp. 1169–1188, 2022, doi: 10.1007/s00170-022-08761-9.
[16] K. L. Li, H. K. Huang, S. F. Tian, and W. Xu, “Improving one-class SVM for anomaly detection,” International Conference on
Machine Learning and Cybernetics, vol. 5, pp. 3077–3081, 2003, doi: 10.1109/icmlc.2003.1260106.
[17] M. Sabokrou, M. Khalooei, M. Fathy, and E. Adeli, “Adversarially learned one-class classifier for novelty detection,” in
Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2018, pp. 3379–3388,
doi: 10.1109/CVPR.2018.00356.
[18] W. Liu et al., “Towards visually explaining variational autoencoders,” in Proceedings of the IEEE Computer Society Conference
on Computer Vision and Pattern Recognition, Nov. 2020, pp. 8642–8651, doi: 10.1109/CVPR42600.2020.00867.
[19] L. Wang, D. Zhang, J. Guo, and Y. Han, “Image anomaly detection using normal data only by latent space resampling,” Applied
Sciences (Switzerland), vol. 10, no. 23, pp. 1–19, 2020, doi: 10.3390/app10238660.
[20] D. M. J. Tax and R. P. W. Duin, “Support vector data description,” Machine Learning, vol. 54, no. 1, pp. 45–66, 2004,
doi: 10.1023/B:MACH.0000008084.60811.49.
[21] S. S. Khan and M. G. Madden, “One-class classification: taxonomy of study and review of techniques,” Knowledge Engineering
Review, vol. 29, no. 3, pp. 345–374, 2014, doi: 10.1017/S026988891300043X.
[22] J. Luo, L. Ding, Z. Pan, G. Ni, and G. Hu, “Research on cost-sensitive learning in one-class anomaly detection algorithms,” in
International Conference on Autonomic and Trusted Computing, 2007, pp. 259–268, doi: 10.1007/978-3-540-73547-2_27.
[23] B. Liu, Y. Xiao, L. Cao, Z. Hao, and F. Deng, “SVDD-based outlier detection on uncertain data,” Knowledge and Information
Systems, vol. 34, no. 3, pp. 597–618, 2013, doi: 10.1007/s10115-012-0484-y.
[24] P. Shi, G. Li, Y. Yuan, and L. Kuang, “Outlier detection using improved support vector data description in wireless sensor
networks,” Sensors (Switzerland), vol. 19, no. 21, 2019, doi: 10.3390/s19214712.
[25] S. M. Erfani, S. Rajasegarar, S. Karunasekera, and C. Leckie, “High-dimensional and large-scale anomaly detection using a linear
one-class SVM with deep learning,” Pattern Recognition, vol. 58, pp. 121–134, 2016, doi: 10.1016/j.patcog.2016.03.028.
[26] M. Maggipinto, C. Masiero, A. Beghi, and G. A. Susto, “A convolutional autoencoder approach for feature extraction in virtual
metrology,” Procedia Manufacturing, vol. 17, pp. 126–133, 2018, doi: 10.1016/j.promfg.2018.10.023.
[27] M. D. Zeiler and R. Fergus, “Visualizing and understanding convolutional networks,” in European conference on computer
vision, 2014, pp. 818–833, doi: 10.1007/978-3-319-10590-1_53.
[28] H. Abdi and L. J. Williams, “Principal component analysis,” Wiley Interdisciplinary Reviews: Computational Statistics, vol. 2,
no. 4, pp. 433–459, 2010, doi: 10.1002/wics.101.
[29] A. Bounsiar and M. G. Madden, “Kernels for one-class support vector machines,” ICISA 2014 - 2014 5th International
Conference on Information Science and Applications, 2014, doi: 10.1109/ICISA.2014.6847419.
[30] P. Bergmann, M. Fauser, D. Sattlegger, and C. Steger, “MVTEC ad-a comprehensive real-world dataset for unsupervised
anomaly detection,” Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition,
vol. 2019-June, pp. 9584–9592, 2019, doi: 10.1109/CVPR.2019.00982.
[31] W.-C. Chang, C.-P. Lee, and C.-J. Lin, “A revisit to support vector data description,” Department of Computer Science, National
Taiwan University, Taipei, Taiwan, Technical Report, 2013.
[32] P. Bergmann, S. Löwe, M. Fauser, D. Sattlegger, and C. Steger, “Improving unsupervised defect segmentation by applying structural
similarity to autoencoders,” in VISIGRAPP 2019 - Proceedings of the 14th International Joint Conference on Computer Vision,
Imaging and Computer Graphics Theory and Applications, Jul. 2019, vol. 5, pp. 372–380, doi: 10.5220/0007364503720380.
9. ISSN: 2252-8938
Int J Artif Intell, Vol. 12, No. 2, June 2023: 912-920
920
BIOGRAPHIES OF AUTHORS
Meryem Chaabi pursed Master degree in industrial engineering from faculty of
Science and Technology, Hassan First University, Morocco. She is currently pursuing PhD at
CCPS Laboratory ENSAM, University of Hassan II, Morocco. She can be contacted at email:
meryem.chaabi-etu@etu.univh2c.ma.
Dr. Mohamed Hamlich is Professor at ENSAM Casablanca, since 2014. In 2013,
he obtained his doctoral thesis in Computer Science from Hassan II University in Casablanca.
His areas of research interests include robotics, artificial intelligence, IoT and big data. He is the
president of the association of Connected Objects and Intelligent Systems and the director of
“CCPS” Laboratory at ENSAM-Casablanca. http://hamlich.sadasc.net/. He can be contacted at
email: moha.hamlich@gmail.com.
Moncef Garouani received his Master’s degree in computer engineering from the
USMBA University of Morocco in 2019. He is currently working towards his PhD degree at the
ULCO university in joint supervision thesis with the UH2C university and the School of
engineering's and business' sciences and technics. His main research topics are machine learning,
AutoML, meta-learning, data mining and explainable AI. He can be contacted at email:
mgarouani@gmail.com.