Pub Date : 2019-09-01DOI: 10.1109/ICICT46931.2019.8977707
Rahul, Himanshu Bansal, Monika
Sentiment analysis uses data mining methods to extract information and data from the web through natural language processing. This consists of emotion artificial intelligent and text analysis. It basically helps in finding out the polarity of word data which is categorized into negative, positive and neutral. Sentiment extraction from data sources is a difficult task because some data sources may have unstructured format of data. In this review paper, we tried to summarize a number of classification techniques used in sentiment analysis stating some of their advantages and disadvantages, performance and their accuracy.In this paper, the various data mining techniques used for the prediction of the heart disease are discussed. With the help of data mining, it is very easy task to make expert system where this plays an important role in the prediction of the health related problems. This helps in solving threat of heart related issues also. Data mining is the extraction of hidden predictive information from large databases which creates enhanced knowledge in the field of pharmaceutical science which helps to predict heart disease. Various data mining techniques are applied here. It produces fast, straightforward assessment of the distinct prediction prototype with the help of Artificial Intelligent techniques.
{"title":"Classification Techniques Used in Sentiment Analysis & Prediction of Heart Disease using Data Mining Techniques: Review","authors":"Rahul, Himanshu Bansal, Monika","doi":"10.1109/ICICT46931.2019.8977707","DOIUrl":"https://doi.org/10.1109/ICICT46931.2019.8977707","url":null,"abstract":"Sentiment analysis uses data mining methods to extract information and data from the web through natural language processing. This consists of emotion artificial intelligent and text analysis. It basically helps in finding out the polarity of word data which is categorized into negative, positive and neutral. Sentiment extraction from data sources is a difficult task because some data sources may have unstructured format of data. In this review paper, we tried to summarize a number of classification techniques used in sentiment analysis stating some of their advantages and disadvantages, performance and their accuracy.In this paper, the various data mining techniques used for the prediction of the heart disease are discussed. With the help of data mining, it is very easy task to make expert system where this plays an important role in the prediction of the health related problems. This helps in solving threat of heart related issues also. Data mining is the extraction of hidden predictive information from large databases which creates enhanced knowledge in the field of pharmaceutical science which helps to predict heart disease. Various data mining techniques are applied here. It produces fast, straightforward assessment of the distinct prediction prototype with the help of Artificial Intelligent techniques.","PeriodicalId":412668,"journal":{"name":"2019 International Conference on Issues and Challenges in Intelligent Computing Techniques (ICICT)","volume":"91 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122539009","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Currently, a lot of research is going in the field of sign language recognition. Recognition of gesture poses a serious challenge to the system due to inconsistent illuminance and background conditions, different skin colours of the hand and each person has his/her own trait of making the gesture. It gets even more difficult with Two Hand Indian Sign Language (THISL) due to the representation of gesture with both hands. There is no proper THISL dataset available to the public. So, we present a THISL dataset consisting of 26 gestures each representing the English alphabet. This dataset consists of 50x50 images of total 9100 in which each gesture is made of 350 images and it is divided into two parts, training and test. The training set consists of 7020 images and the test set consists of 2080 images. In this paper, THISL dataset is validated on various classification models of machine learning and overall accuracy of 91.72% is achieved. This dataset serves a very good purpose for benchmarking machine learning algorithms and it is freely available to people on request to authors.
{"title":"Two Hand Indian Sign Language dataset for benchmarking classification models of Machine Learning","authors":"Leela Surya Teja Mangamuri, Lakshay Jain, Abhishek Sharmay","doi":"10.1109/ICICT46931.2019.8977713","DOIUrl":"https://doi.org/10.1109/ICICT46931.2019.8977713","url":null,"abstract":"Currently, a lot of research is going in the field of sign language recognition. Recognition of gesture poses a serious challenge to the system due to inconsistent illuminance and background conditions, different skin colours of the hand and each person has his/her own trait of making the gesture. It gets even more difficult with Two Hand Indian Sign Language (THISL) due to the representation of gesture with both hands. There is no proper THISL dataset available to the public. So, we present a THISL dataset consisting of 26 gestures each representing the English alphabet. This dataset consists of 50x50 images of total 9100 in which each gesture is made of 350 images and it is divided into two parts, training and test. The training set consists of 7020 images and the test set consists of 2080 images. In this paper, THISL dataset is validated on various classification models of machine learning and overall accuracy of 91.72% is achieved. This dataset serves a very good purpose for benchmarking machine learning algorithms and it is freely available to people on request to authors.","PeriodicalId":412668,"journal":{"name":"2019 International Conference on Issues and Challenges in Intelligent Computing Techniques (ICICT)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122734021","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/ICICT46931.2019.8977642
Monika Mathur, Vivek Upadhyaya, Rahul Srivastava
An algorithm based on Stockwell Transform focused on processing of communication signals to detect harmonics and transient disturbances superimposed on the signals is presented in this paper. These disturbances are being superimposed on the signals in the communication channel or at the transmitter or the receiver stations. Investigated transient disturbances include impulsive transient and oscillatory transients. Communication signals incorporating harmonics or transient disturbance are decomposed with the help of Stockwell Transform and S-matrix is derived. A summation of absolute values curve, median curve and maximum absolute values plot are proposed to detect disturbances. These curves are obtained from S-matrix. On comparing these plots of signal having harmonics or transient disturbances with respective curves of pure sinusoidal communication signal, superimposed harmonics or transient disturbance have been detected successfully. Effectiveness of the proposed approach is established using the MATLAB software.
{"title":"Algorithm Based on Stockwell Transform for Processing of Communication Signal to Detect Superimposed Harmonics and Transient Disturbances","authors":"Monika Mathur, Vivek Upadhyaya, Rahul Srivastava","doi":"10.1109/ICICT46931.2019.8977642","DOIUrl":"https://doi.org/10.1109/ICICT46931.2019.8977642","url":null,"abstract":"An algorithm based on Stockwell Transform focused on processing of communication signals to detect harmonics and transient disturbances superimposed on the signals is presented in this paper. These disturbances are being superimposed on the signals in the communication channel or at the transmitter or the receiver stations. Investigated transient disturbances include impulsive transient and oscillatory transients. Communication signals incorporating harmonics or transient disturbance are decomposed with the help of Stockwell Transform and S-matrix is derived. A summation of absolute values curve, median curve and maximum absolute values plot are proposed to detect disturbances. These curves are obtained from S-matrix. On comparing these plots of signal having harmonics or transient disturbances with respective curves of pure sinusoidal communication signal, superimposed harmonics or transient disturbance have been detected successfully. Effectiveness of the proposed approach is established using the MATLAB software.","PeriodicalId":412668,"journal":{"name":"2019 International Conference on Issues and Challenges in Intelligent Computing Techniques (ICICT)","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132605663","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/ICICT46931.2019.8977703
Neelam Rawat, J. Sodhi, R. Tyagi
Detection of moving object is a challenging task for any video surveillance. Generally, a video surveillance consists of three levels of processing – moving object extraction, recognition and tracking – that further process for decision of the corresponding activities. As pre- and post-processing of the video might be necessary to improve the detection of moving objects, we have proposed a method that uses the coordinates of the boundary of an object with a thick white line. This paper basically shows the results through which we specify the initial search condition, the connectivity and how many pixels should be returned and the direction in which to perform the search. Further traces the exterior boundaries of objects.
{"title":"Acquiring and Analyzing Movement Detection through Image Granulation","authors":"Neelam Rawat, J. Sodhi, R. Tyagi","doi":"10.1109/ICICT46931.2019.8977703","DOIUrl":"https://doi.org/10.1109/ICICT46931.2019.8977703","url":null,"abstract":"Detection of moving object is a challenging task for any video surveillance. Generally, a video surveillance consists of three levels of processing – moving object extraction, recognition and tracking – that further process for decision of the corresponding activities. As pre- and post-processing of the video might be necessary to improve the detection of moving objects, we have proposed a method that uses the coordinates of the boundary of an object with a thick white line. This paper basically shows the results through which we specify the initial search condition, the connectivity and how many pixels should be returned and the direction in which to perform the search. Further traces the exterior boundaries of objects.","PeriodicalId":412668,"journal":{"name":"2019 International Conference on Issues and Challenges in Intelligent Computing Techniques (ICICT)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115528150","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/ICICT46931.2019.8977657
Jayesh R Solanki, Divykant Meva
The world’s largest democracy has adopted electoral reforms in Assembly as well as Parliamentary elections. Electronic Voting Machine (EVM) has replaced the paper ballot system. However, there are serious concerns being raised with regards to the credibility and reliability of the EVM’s. This resulted in Voter Verified Paper Audit Trail (VVPAT) being attached to the EVM, which was found to be unverifiable and non-auditable. The primary focus of this paper is to provide a comparison between Ballot Paper Voting System and EVM highlighting the various challenges of the existing electoral system. This paper provides an insight into Block chain technology, its impact and revolution it can bring in the field of Electoral reforms in India. The comparison parameters like Time, Cost, Transparency, Risk factor, Verification/Auditing process, Authentication etc. are considered for evaluation of the different methodology discussed.
{"title":"Comparative Study Indian Electoral Reforms in Indian Context","authors":"Jayesh R Solanki, Divykant Meva","doi":"10.1109/ICICT46931.2019.8977657","DOIUrl":"https://doi.org/10.1109/ICICT46931.2019.8977657","url":null,"abstract":"The world’s largest democracy has adopted electoral reforms in Assembly as well as Parliamentary elections. Electronic Voting Machine (EVM) has replaced the paper ballot system. However, there are serious concerns being raised with regards to the credibility and reliability of the EVM’s. This resulted in Voter Verified Paper Audit Trail (VVPAT) being attached to the EVM, which was found to be unverifiable and non-auditable. The primary focus of this paper is to provide a comparison between Ballot Paper Voting System and EVM highlighting the various challenges of the existing electoral system. This paper provides an insight into Block chain technology, its impact and revolution it can bring in the field of Electoral reforms in India. The comparison parameters like Time, Cost, Transparency, Risk factor, Verification/Auditing process, Authentication etc. are considered for evaluation of the different methodology discussed.","PeriodicalId":412668,"journal":{"name":"2019 International Conference on Issues and Challenges in Intelligent Computing Techniques (ICICT)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128927448","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/ICICT46931.2019.8977709
Sunita Chand, V. P. Vishwakarma
Leukemia is a fatal disease that is commonly found in children and also in adults above 55 years of age. It is also known as cancer of blood or bone marrow. [1] It can be categorized into myelogenous leukemia or lymphocytic on the basis of the cells affected by the disease. As the symptoms of the disease are very common like fever, fatigue and body ache, it is not easily detectable at early stages which prove fatal at later stages. So diagnosing it at early stage is crucial for the better prognosis of disease. The paper presents a comparative analysis of extensively used machine learning (ML) algorithm SVM and the relatively new ML algorithm i.e., extreme learning machine for predicting Leukemia. The classification is based on the segmentation of blood smear images publically available dataset ALL-IDB1. The results shows that ELM with an accuracy of 92.2448% outperforms SVM with accuracy 86.3636%.
{"title":"Leukemia Diagnosis using Computational Intelligence","authors":"Sunita Chand, V. P. Vishwakarma","doi":"10.1109/ICICT46931.2019.8977709","DOIUrl":"https://doi.org/10.1109/ICICT46931.2019.8977709","url":null,"abstract":"Leukemia is a fatal disease that is commonly found in children and also in adults above 55 years of age. It is also known as cancer of blood or bone marrow. [1] It can be categorized into myelogenous leukemia or lymphocytic on the basis of the cells affected by the disease. As the symptoms of the disease are very common like fever, fatigue and body ache, it is not easily detectable at early stages which prove fatal at later stages. So diagnosing it at early stage is crucial for the better prognosis of disease. The paper presents a comparative analysis of extensively used machine learning (ML) algorithm SVM and the relatively new ML algorithm i.e., extreme learning machine for predicting Leukemia. The classification is based on the segmentation of blood smear images publically available dataset ALL-IDB1. The results shows that ELM with an accuracy of 92.2448% outperforms SVM with accuracy 86.3636%.","PeriodicalId":412668,"journal":{"name":"2019 International Conference on Issues and Challenges in Intelligent Computing Techniques (ICICT)","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128928205","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/ICICT46931.2019.8977644
Deepak Kumar
For the security related issues over internet two main techniques are used first is Cryptography and second is Steganography. Both are basically used for data security. Cryptography transforms the data from one form to another form while steganography hide data in an image in such a way that it cannot be detected by human eyes. This paper introduces an image steganography method using YCbCr color model based on Least Significant Bit (LSB). In this paper proposed method convert the image from RGB to YCbCr color space then secret data is hidden inside YCbCr color space using least significant bit and after hiding the data, convert it back to RGB color space. The proposed technique is evaluated by objective analysis. Different techniques of cryptography are compared using Mean Square Error (MSE) and Peak Signal to Noise Ratio (PSNR). It is observed that the proposed method has high PSNR and low MSE which shows the proposed approach is very efficient to hide data in an image
{"title":"Hiding Text In Color Image Using YCbCr Color Model: An Image Steganography approach","authors":"Deepak Kumar","doi":"10.1109/ICICT46931.2019.8977644","DOIUrl":"https://doi.org/10.1109/ICICT46931.2019.8977644","url":null,"abstract":"For the security related issues over internet two main techniques are used first is Cryptography and second is Steganography. Both are basically used for data security. Cryptography transforms the data from one form to another form while steganography hide data in an image in such a way that it cannot be detected by human eyes. This paper introduces an image steganography method using YCbCr color model based on Least Significant Bit (LSB). In this paper proposed method convert the image from RGB to YCbCr color space then secret data is hidden inside YCbCr color space using least significant bit and after hiding the data, convert it back to RGB color space. The proposed technique is evaluated by objective analysis. Different techniques of cryptography are compared using Mean Square Error (MSE) and Peak Signal to Noise Ratio (PSNR). It is observed that the proposed method has high PSNR and low MSE which shows the proposed approach is very efficient to hide data in an image","PeriodicalId":412668,"journal":{"name":"2019 International Conference on Issues and Challenges in Intelligent Computing Techniques (ICICT)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127927181","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/ICICT46931.2019.8977697
Kanica Sachdev, M. Gupta
The identification of drug protein associations assists the exploration of novel drugs, drug repurposing and drug side effect identification. The experimental evaluation of these interactions requires extensive capital and money. Thus, in-silico computational methods are being developed to aid the interaction prediction. These techniques have been broadly grouped into similarity based approaches and feature based approaches. This paper proposes a novel feature based approach to identify the probable drug protein communications. The method is based on Support Vector Machine classifier. Support Vector Machines have shown a satisfactory performance in many applications related to the pharmacology domain. To further improve the accuracy and reduce the computational complexity, dimensionality reduction by PCA has been proposed. The proposed technique achieves an AUC score of 0.822. The method has been compared to various other state of the art methods based on their respective AUC scores. The comparison has shown that the proposed approach has a better performance in contrast to the other techniques.
{"title":"An Improved Approach for Predicting Drug Target Interactions","authors":"Kanica Sachdev, M. Gupta","doi":"10.1109/ICICT46931.2019.8977697","DOIUrl":"https://doi.org/10.1109/ICICT46931.2019.8977697","url":null,"abstract":"The identification of drug protein associations assists the exploration of novel drugs, drug repurposing and drug side effect identification. The experimental evaluation of these interactions requires extensive capital and money. Thus, in-silico computational methods are being developed to aid the interaction prediction. These techniques have been broadly grouped into similarity based approaches and feature based approaches. This paper proposes a novel feature based approach to identify the probable drug protein communications. The method is based on Support Vector Machine classifier. Support Vector Machines have shown a satisfactory performance in many applications related to the pharmacology domain. To further improve the accuracy and reduce the computational complexity, dimensionality reduction by PCA has been proposed. The proposed technique achieves an AUC score of 0.822. The method has been compared to various other state of the art methods based on their respective AUC scores. The comparison has shown that the proposed approach has a better performance in contrast to the other techniques.","PeriodicalId":412668,"journal":{"name":"2019 International Conference on Issues and Challenges in Intelligent Computing Techniques (ICICT)","volume":"407 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121005750","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/ICICT46931.2019.8977650
Arjun Puri, M. Gupta
Classification is mainly challenged by the problems in the dataset. When dataset have uneven distribution of data among classes, then class imbalance problem arise. Class imbalance with noise creates immense effect on classification of instances of classes. The main focus of this article is to provide the detail comparative analysis of seven Resampling techniques under 16 noisy imbalanced datasets using C4.5 classifier. The performance evaluation is done by using AUC, F1 score, G-mean. Based on the evaluation, article inferred that SMOTE-ENN perform better than rest of resampling techniques.
{"title":"Comparative Analysis of Resampling Techniques under Noisy Imbalanced Datasets","authors":"Arjun Puri, M. Gupta","doi":"10.1109/ICICT46931.2019.8977650","DOIUrl":"https://doi.org/10.1109/ICICT46931.2019.8977650","url":null,"abstract":"Classification is mainly challenged by the problems in the dataset. When dataset have uneven distribution of data among classes, then class imbalance problem arise. Class imbalance with noise creates immense effect on classification of instances of classes. The main focus of this article is to provide the detail comparative analysis of seven Resampling techniques under 16 noisy imbalanced datasets using C4.5 classifier. The performance evaluation is done by using AUC, F1 score, G-mean. Based on the evaluation, article inferred that SMOTE-ENN perform better than rest of resampling techniques.","PeriodicalId":412668,"journal":{"name":"2019 International Conference on Issues and Challenges in Intelligent Computing Techniques (ICICT)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121502388","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-09-01DOI: 10.1109/ICICT46931.2019.8977696
Somit Mittal, Chahes Chopra, A. Trivedi, P. Chanak
Surface inspection is one of the most challenging tasks in the manufacturing industry. Defect classification and segmentation are the two main tasks associated with surface inspection. The major challenge lies in the collection of the dataset as it is a very costly procedure and the occurrences of defected samples are very less as compared to non defective samples. Therefore, it becomes important to devise a method that can leverage the limited data available and can also handle the class imbalance between the defected and non defected samples. In this paper, a deep learning approach is proposed that uses pertained networks to perform defect segmentation on industrial surfaces. The deep learning approach consists of an encoder and decoder architecture where on the encoder side, VGG is used with pertained imagenet weights for faster training of the model and on the decoder side, the UNet decoder model is used. The evaluation of the approach shows that the proposed method can be used for surface inspection in various industrial applications.
{"title":"Defect Segmentation in Surfaces using Deep Learning","authors":"Somit Mittal, Chahes Chopra, A. Trivedi, P. Chanak","doi":"10.1109/ICICT46931.2019.8977696","DOIUrl":"https://doi.org/10.1109/ICICT46931.2019.8977696","url":null,"abstract":"Surface inspection is one of the most challenging tasks in the manufacturing industry. Defect classification and segmentation are the two main tasks associated with surface inspection. The major challenge lies in the collection of the dataset as it is a very costly procedure and the occurrences of defected samples are very less as compared to non defective samples. Therefore, it becomes important to devise a method that can leverage the limited data available and can also handle the class imbalance between the defected and non defected samples. In this paper, a deep learning approach is proposed that uses pertained networks to perform defect segmentation on industrial surfaces. The deep learning approach consists of an encoder and decoder architecture where on the encoder side, VGG is used with pertained imagenet weights for faster training of the model and on the decoder side, the UNet decoder model is used. The evaluation of the approach shows that the proposed method can be used for surface inspection in various industrial applications.","PeriodicalId":412668,"journal":{"name":"2019 International Conference on Issues and Challenges in Intelligent Computing Techniques (ICICT)","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126800615","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：