Pub Date : 2023-03-17DOI: 10.1109/iCoMET57998.2023.10099334
Muhammad Jawad, Yu Haitao, Zahoor Ahmad, Basharat Ullah, Baheej Alghamdi
This article examines a new linear oscillating actuator (LOA) design that uses rectangular-shaped core materials and permanent magnets (PMs). The paper's primary objective is to examine a novel LOA topology with a rectangular-shaped core and PMs as an alternative to the tubular LOA. A Static core material is housed between the mover's components, and two stator cores with two coils each are placed on either side of the mover. The dimensions of all the parameters are swept for optimization, and the optimal parameter dimension is chosen based on the optimal value of the electromagnetic (EM) force. An analysis is done for output parameters like EM force and stroke. EM force per PM mass of the investigated design of LOA is 67 percent higher than of conventional rectangular-shaped LOA. Additionally, the proposed design's EM force density is 23.8 percent higher than that of the conventional design of LOA. Furthermore, the stroke of the proposed LOA is also feasible and more than most of the designs of the LOA. Additionally, the proposed design of the LOA is simple structure, low-cost, and feasible for fabrication.
{"title":"Design Optimization and Performance Analysis of Rectangular Structured Moving Magnet Linear Actuator","authors":"Muhammad Jawad, Yu Haitao, Zahoor Ahmad, Basharat Ullah, Baheej Alghamdi","doi":"10.1109/iCoMET57998.2023.10099334","DOIUrl":"https://doi.org/10.1109/iCoMET57998.2023.10099334","url":null,"abstract":"This article examines a new linear oscillating actuator (LOA) design that uses rectangular-shaped core materials and permanent magnets (PMs). The paper's primary objective is to examine a novel LOA topology with a rectangular-shaped core and PMs as an alternative to the tubular LOA. A Static core material is housed between the mover's components, and two stator cores with two coils each are placed on either side of the mover. The dimensions of all the parameters are swept for optimization, and the optimal parameter dimension is chosen based on the optimal value of the electromagnetic (EM) force. An analysis is done for output parameters like EM force and stroke. EM force per PM mass of the investigated design of LOA is 67 percent higher than of conventional rectangular-shaped LOA. Additionally, the proposed design's EM force density is 23.8 percent higher than that of the conventional design of LOA. Furthermore, the stroke of the proposed LOA is also feasible and more than most of the designs of the LOA. Additionally, the proposed design of the LOA is simple structure, low-cost, and feasible for fabrication.","PeriodicalId":369792,"journal":{"name":"2023 4th International Conference on Computing, Mathematics and Engineering Technologies (iCoMET)","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115964147","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 : 2023-03-17DOI: 10.1109/iCoMET57998.2023.10099282
Rafiq Asghar, Muhammad Junaid Anwar, Hamid Wadood, Haider Saleem, Nauman Rasul, Z. Ullah
The energy sector is one of the primary promising renewable energy sources. To meet consumer demand, more and more energy generation units are required. From the development of generating stations to reasonable running conditions, the responsible authorities do all the analysis and measures. Most of these generating stations depend on fossil fuels, creating constant problems for society, such as the production of Greenhouse Gases (GHG), which depletes the ozone layer and makes the country economically down. Contrarily, Renewable Energy Sources (RES) for energy production are economical and user-friendly. Wind Energy Source (WES) is one of the major sources of energy. The paper presents the extensively analyzed benefits of WES from all perspectives.
{"title":"Promising Features of Wind Energy: A Glance Overview","authors":"Rafiq Asghar, Muhammad Junaid Anwar, Hamid Wadood, Haider Saleem, Nauman Rasul, Z. Ullah","doi":"10.1109/iCoMET57998.2023.10099282","DOIUrl":"https://doi.org/10.1109/iCoMET57998.2023.10099282","url":null,"abstract":"The energy sector is one of the primary promising renewable energy sources. To meet consumer demand, more and more energy generation units are required. From the development of generating stations to reasonable running conditions, the responsible authorities do all the analysis and measures. Most of these generating stations depend on fossil fuels, creating constant problems for society, such as the production of Greenhouse Gases (GHG), which depletes the ozone layer and makes the country economically down. Contrarily, Renewable Energy Sources (RES) for energy production are economical and user-friendly. Wind Energy Source (WES) is one of the major sources of energy. The paper presents the extensively analyzed benefits of WES from all perspectives.","PeriodicalId":369792,"journal":{"name":"2023 4th International Conference on Computing, Mathematics and Engineering Technologies (iCoMET)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115457703","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 : 2023-03-17DOI: 10.1109/iCoMET57998.2023.10099343
Subhan Ali, A. Imran, Zenun Kastrati, Sher Muhammad Daudpota
Understanding complex machine learning and artificial intelligence models have always been challenging because these models are black-box, and often we don't know what information models rely upon to infer. Explainable Artificial Intelligence (XAI) has emerged as a new exciting field to explain and understand these machine learning models as humans can understand and improve them. In the past few years, there have been numerous research articles on explainable artificial intelligence for medical and healthcare. 1687 documents are being studied and analysed using bibliometric methods in this work. There are certain systematic reviews on the same topic, but this study is the first of its kind to use a quantitative method to analyze a large number of publications. The results of this study show that the research in this field took pace in 2011, and there have been quite many publications in the following years. We have also identified top-cited journals and articles. Through thematic analysis, we have found some important thematic areas of research in the field of XAI for medical and healthcare. The findings showed that the USA is the global leader in XAI research, followed by China and Canada at second and third place, respectively.
{"title":"Visualizing Research on Explainable Artificial Intelligence for Medical and Healthcare","authors":"Subhan Ali, A. Imran, Zenun Kastrati, Sher Muhammad Daudpota","doi":"10.1109/iCoMET57998.2023.10099343","DOIUrl":"https://doi.org/10.1109/iCoMET57998.2023.10099343","url":null,"abstract":"Understanding complex machine learning and artificial intelligence models have always been challenging because these models are black-box, and often we don't know what information models rely upon to infer. Explainable Artificial Intelligence (XAI) has emerged as a new exciting field to explain and understand these machine learning models as humans can understand and improve them. In the past few years, there have been numerous research articles on explainable artificial intelligence for medical and healthcare. 1687 documents are being studied and analysed using bibliometric methods in this work. There are certain systematic reviews on the same topic, but this study is the first of its kind to use a quantitative method to analyze a large number of publications. The results of this study show that the research in this field took pace in 2011, and there have been quite many publications in the following years. We have also identified top-cited journals and articles. Through thematic analysis, we have found some important thematic areas of research in the field of XAI for medical and healthcare. The findings showed that the USA is the global leader in XAI research, followed by China and Canada at second and third place, respectively.","PeriodicalId":369792,"journal":{"name":"2023 4th International Conference on Computing, Mathematics and Engineering Technologies (iCoMET)","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116036016","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 : 2023-03-17DOI: 10.1109/iCoMET57998.2023.10099119
Muneeba Humayoun, Hana Sharif, Faisal Rehman, Shahbaz Shaukat, Muhbat Ullah, Hadia Maqsood, C. Ali, Razia Iftikhar, Adil Hussain Chandio
Due to the large number of things and information devices, not all IoT applications can be satisfied by processing data in the cloud. Due to the cloud's constrained ability to process and share data, edge computing, or the act of initiating IoT edge data processing and connected devices' transformation from intelligent devices to gadgets, was developed. Machine learning is the key instrument. It is important to include information inference as a continuum in the cloud-to-things approach. Reviewing machine functions that are connected to the Internet, from the cloud all the way down to embedded devices. Many uses for machines learning to handle application data management and processing responsibilities are examined. The most current machine learning apps for IoT are gathered, and they all agree on their feedback and application space. The type of data, the machine learning methods used, and the locations belong to the continuum from clouds to objects. The issues and future directions of IoT machine learning research are spoken about. Additionally, employing methods for categorization using machine learning, papers on “machine” learning in IoT are meticulously retrieved and reviewed. Next, with the expansion of recognized subjects and application domains, difficulties and search are moving in the direction of effective machine learning for the IoT. In addition, articles on the IoT's “machine” learning are painstakingly retrieved, then classified using machine learning methods.
{"title":"From Cloud Down to Things: An Overview of Machine Learning in Internet of Things","authors":"Muneeba Humayoun, Hana Sharif, Faisal Rehman, Shahbaz Shaukat, Muhbat Ullah, Hadia Maqsood, C. Ali, Razia Iftikhar, Adil Hussain Chandio","doi":"10.1109/iCoMET57998.2023.10099119","DOIUrl":"https://doi.org/10.1109/iCoMET57998.2023.10099119","url":null,"abstract":"Due to the large number of things and information devices, not all IoT applications can be satisfied by processing data in the cloud. Due to the cloud's constrained ability to process and share data, edge computing, or the act of initiating IoT edge data processing and connected devices' transformation from intelligent devices to gadgets, was developed. Machine learning is the key instrument. It is important to include information inference as a continuum in the cloud-to-things approach. Reviewing machine functions that are connected to the Internet, from the cloud all the way down to embedded devices. Many uses for machines learning to handle application data management and processing responsibilities are examined. The most current machine learning apps for IoT are gathered, and they all agree on their feedback and application space. The type of data, the machine learning methods used, and the locations belong to the continuum from clouds to objects. The issues and future directions of IoT machine learning research are spoken about. Additionally, employing methods for categorization using machine learning, papers on “machine” learning in IoT are meticulously retrieved and reviewed. Next, with the expansion of recognized subjects and application domains, difficulties and search are moving in the direction of effective machine learning for the IoT. In addition, articles on the IoT's “machine” learning are painstakingly retrieved, then classified using machine learning methods.","PeriodicalId":369792,"journal":{"name":"2023 4th International Conference on Computing, Mathematics and Engineering Technologies (iCoMET)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126906249","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 : 2023-03-17DOI: 10.1109/iCoMET57998.2023.10099146
Naseer Ahmad, Tahsinullah, Shoaib Ahmed, Haris Shahbaz, Surat Khan
Research on flux reversal machine is taking importance day by day because of its unique characteristics. This research through light over different performance of flux reversal machine like three-phase flux linkage, cogging torque, back EMF, average torque, self, and mutual inductance. Moreover, average torque vs stack length and torque ripple is analyzed at different rotor pole and noticed a tremendous shift in performance with changing the number of magnet pairs per pole and observed that two-pairs (4-PM) of PM per pole gives the best result then a pair (2-PM) of PM per pole. All the analysis is done under the same condition, parameters, and materials. As a result, it is concluded that 10-pole model gives higher torque density and better average torque, but with high spicks then 11,13 and 17-pole model. In terms of average torque, the 10-pole model is dominant overall with 3.30Nm (17.097%) higher than 13-pole model, followed by 11-pole model with average torque value of 2.656Nm, while the 17-pole model produce the lowest average torque but spike less. At the end average torque is briefly discussed with respect to current density and stack length.
{"title":"Performance Analysis of Outer Rotor Flux Reversal Machine at Different Rotor Poles","authors":"Naseer Ahmad, Tahsinullah, Shoaib Ahmed, Haris Shahbaz, Surat Khan","doi":"10.1109/iCoMET57998.2023.10099146","DOIUrl":"https://doi.org/10.1109/iCoMET57998.2023.10099146","url":null,"abstract":"Research on flux reversal machine is taking importance day by day because of its unique characteristics. This research through light over different performance of flux reversal machine like three-phase flux linkage, cogging torque, back EMF, average torque, self, and mutual inductance. Moreover, average torque vs stack length and torque ripple is analyzed at different rotor pole and noticed a tremendous shift in performance with changing the number of magnet pairs per pole and observed that two-pairs (4-PM) of PM per pole gives the best result then a pair (2-PM) of PM per pole. All the analysis is done under the same condition, parameters, and materials. As a result, it is concluded that 10-pole model gives higher torque density and better average torque, but with high spicks then 11,13 and 17-pole model. In terms of average torque, the 10-pole model is dominant overall with 3.30Nm (17.097%) higher than 13-pole model, followed by 11-pole model with average torque value of 2.656Nm, while the 17-pole model produce the lowest average torque but spike less. At the end average torque is briefly discussed with respect to current density and stack length.","PeriodicalId":369792,"journal":{"name":"2023 4th International Conference on Computing, Mathematics and Engineering Technologies (iCoMET)","volume":"88 15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126308905","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 : 2023-03-17DOI: 10.1109/iCoMET57998.2023.10099364
Shylendra Kumar, R. Kumar, Saad
This research paper presents a real-time detection of road-based objects using SSD MobileNet-v2 FPNlite. This model uses the Single Shot Detector (SSD) architecture with MobileNet-v2 as the backbone and Feature Pyramid Network lite (FPNlite) as the feature extractor. This approach combines the advantages of both SSD and MobileNet-v2 for object detection while maintaining low computational complexity. In order to evaluate the performance of the model, a new benchmark dataset is explicitly created for this study, which includes a wide range of images captured from various sources such as cameras mounted on vehicles and street-level cameras. The dataset contains a diverse set of objects and scenes, making it suitable for testing the robustness and generalization ability of the system. The results of the experiments demonstrate the effectiveness of the model. In addition, the newly developed benchmark dataset can be used as a reference for further research in the field.
{"title":"Real-Time Detection of Road-Based Objects using SSD MobileNet-v2 FPNlite with a new Benchmark Dataset","authors":"Shylendra Kumar, R. Kumar, Saad","doi":"10.1109/iCoMET57998.2023.10099364","DOIUrl":"https://doi.org/10.1109/iCoMET57998.2023.10099364","url":null,"abstract":"This research paper presents a real-time detection of road-based objects using SSD MobileNet-v2 FPNlite. This model uses the Single Shot Detector (SSD) architecture with MobileNet-v2 as the backbone and Feature Pyramid Network lite (FPNlite) as the feature extractor. This approach combines the advantages of both SSD and MobileNet-v2 for object detection while maintaining low computational complexity. In order to evaluate the performance of the model, a new benchmark dataset is explicitly created for this study, which includes a wide range of images captured from various sources such as cameras mounted on vehicles and street-level cameras. The dataset contains a diverse set of objects and scenes, making it suitable for testing the robustness and generalization ability of the system. The results of the experiments demonstrate the effectiveness of the model. In addition, the newly developed benchmark dataset can be used as a reference for further research in the field.","PeriodicalId":369792,"journal":{"name":"2023 4th International Conference on Computing, Mathematics and Engineering Technologies (iCoMET)","volume":"150 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131017882","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 : 2023-03-17DOI: 10.1109/iCoMET57998.2023.10099171
Rahool Rai, Ali Raza Larik, Kashif Ahmed, Sudhakar Kumaramasy, Asad A. Zaidi
the prospect of our globe is convolutedly tangled with the coming adoptions of energy, effective manipulation of renewable energy cradles is flattering progressively vital for up-to-date world as conventional fuels are perilous to environment and cannot withstand supply for extended period since they are depleting, ultimately they will diminish one day. Moreover, mandated energy is snowballing rapidly. In this scenario, solar energy is being perceived as possible variable resource for ever-growing starvation of the energy for the progress of nation at large and adopted globally. However, the small efficiency and intermittent availability of solar energy has called for the development by different techniques to enhance the productivity of the solar heater (Air) by coating the finned absorber with black paint. Naturally, black color absorbs the maximum heat from the irradiance. Which ultimately escalates the efficiency of SAH in form of solar thermal energy. Results depicted that, having reached 60 minutes of heating through solar radiations, solar air collectors attached with black painted finned absorber reached the 50 % of efficiency of solar irradiation of 900–1000 W/m2.
{"title":"Comparative Analysis of finned absorber plate with and without black paint in Solar Air Heater","authors":"Rahool Rai, Ali Raza Larik, Kashif Ahmed, Sudhakar Kumaramasy, Asad A. Zaidi","doi":"10.1109/iCoMET57998.2023.10099171","DOIUrl":"https://doi.org/10.1109/iCoMET57998.2023.10099171","url":null,"abstract":"the prospect of our globe is convolutedly tangled with the coming adoptions of energy, effective manipulation of renewable energy cradles is flattering progressively vital for up-to-date world as conventional fuels are perilous to environment and cannot withstand supply for extended period since they are depleting, ultimately they will diminish one day. Moreover, mandated energy is snowballing rapidly. In this scenario, solar energy is being perceived as possible variable resource for ever-growing starvation of the energy for the progress of nation at large and adopted globally. However, the small efficiency and intermittent availability of solar energy has called for the development by different techniques to enhance the productivity of the solar heater (Air) by coating the finned absorber with black paint. Naturally, black color absorbs the maximum heat from the irradiance. Which ultimately escalates the efficiency of SAH in form of solar thermal energy. Results depicted that, having reached 60 minutes of heating through solar radiations, solar air collectors attached with black painted finned absorber reached the 50 % of efficiency of solar irradiation of 900–1000 W/m2.","PeriodicalId":369792,"journal":{"name":"2023 4th International Conference on Computing, Mathematics and Engineering Technologies (iCoMET)","volume":"101 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128587407","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 : 2023-03-17DOI: 10.1109/iCoMET57998.2023.10099324
Hana Sharif, Faisal Rehman, Amina Rida, Chaudhry Nouman Ali, Rana Zeeshan Zulfiqar, Salman Akram, Hina Kirn, Ali Hussain, Razia Iftikhar
Artificial Neural Networks (ANN) and deep learning have been instrumental in the advancement of technology around the world for about 50 years, but because of the high costs associated with developing an optimal training and testing dataset, researchers have had to deal with several issues such as segmentation of images with low spatial resolution, object recognition, classification, and their use in the processing of satellite images has yet to reach its full potential. This work includes a thorough assessment of a significant body of research literature as well as the most important publications released in the last decade. The IEEE digital library, Science Direct, and SCOPUS system database indexing repository were the key sources for the review after applying various criteria, 386 publications pertaining to the case study were discovered. With 30 of them, grounds for exclusion and inclusion are discussed in further detail. Finding an upward trend in the level of research done in recent years indicates that this artificial neural network technology is getting more and more popular and curious.
{"title":"Application of Artificial Neural Networks inSatellite Imaging – A Systematic Review","authors":"Hana Sharif, Faisal Rehman, Amina Rida, Chaudhry Nouman Ali, Rana Zeeshan Zulfiqar, Salman Akram, Hina Kirn, Ali Hussain, Razia Iftikhar","doi":"10.1109/iCoMET57998.2023.10099324","DOIUrl":"https://doi.org/10.1109/iCoMET57998.2023.10099324","url":null,"abstract":"Artificial Neural Networks (ANN) and deep learning have been instrumental in the advancement of technology around the world for about 50 years, but because of the high costs associated with developing an optimal training and testing dataset, researchers have had to deal with several issues such as segmentation of images with low spatial resolution, object recognition, classification, and their use in the processing of satellite images has yet to reach its full potential. This work includes a thorough assessment of a significant body of research literature as well as the most important publications released in the last decade. The IEEE digital library, Science Direct, and SCOPUS system database indexing repository were the key sources for the review after applying various criteria, 386 publications pertaining to the case study were discovered. With 30 of them, grounds for exclusion and inclusion are discussed in further detail. Finding an upward trend in the level of research done in recent years indicates that this artificial neural network technology is getting more and more popular and curious.","PeriodicalId":369792,"journal":{"name":"2023 4th International Conference on Computing, Mathematics and Engineering Technologies (iCoMET)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128856205","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 : 2023-03-17DOI: 10.1109/iCoMET57998.2023.10099097
Mumtaz A. Kaloi, Asif Ali, Irfan Ali Babar, K. Mujeeb
Retinopathy detection based on deep learning methods is a challenging problem, especially the diabetic retinopathy (DR) brings so many technical complications in medical image processing. Recently, label smoothing regularization has proved to be a better option to improve the performance of deep learning models. Therefore, in this paper, we introduce a dual-stream multi-task learning model along with a novel weighted label smoothing regularization loss (WLSRL) to detect retinopathy. The proposed model uses a dual-stream network by incorporating separable and conventional convolutional neural networks to detect diabetic retinopathy. The model is designed to classify numerous retinal diseases on two different types of data. The data $Delta_{1}$ is based on stereoscopic fundus photographs and $Delta_{2}$ consists of OCT-based retinal images. The model is trained and tested on both data separately. We perform two classification tasks TF1, TF2 on $Delta_{1}$ and TO1, TO2 on $Delta_{2}$. The task TF1 is for the classification of fundus photographs as normal and abnormal, whereas TF2 is for DR grading. Similarly, the task TO1 classifies OCT-based images into four classes, whereas the task TO2 classifies images as normal and abnormal. The empirical results show that the model achieves competitive results for retinopathy classification and grading using multitask learning with WLSRL.
{"title":"Label Smoothing Loss with Dual-Stream Network Using Separable Convolutional Layers for Retinopathy Grading and Classification","authors":"Mumtaz A. Kaloi, Asif Ali, Irfan Ali Babar, K. Mujeeb","doi":"10.1109/iCoMET57998.2023.10099097","DOIUrl":"https://doi.org/10.1109/iCoMET57998.2023.10099097","url":null,"abstract":"Retinopathy detection based on deep learning methods is a challenging problem, especially the diabetic retinopathy (DR) brings so many technical complications in medical image processing. Recently, label smoothing regularization has proved to be a better option to improve the performance of deep learning models. Therefore, in this paper, we introduce a dual-stream multi-task learning model along with a novel weighted label smoothing regularization loss (WLSRL) to detect retinopathy. The proposed model uses a dual-stream network by incorporating separable and conventional convolutional neural networks to detect diabetic retinopathy. The model is designed to classify numerous retinal diseases on two different types of data. The data $Delta_{1}$ is based on stereoscopic fundus photographs and $Delta_{2}$ consists of OCT-based retinal images. The model is trained and tested on both data separately. We perform two classification tasks TF1, TF2 on $Delta_{1}$ and TO1, TO2 on $Delta_{2}$. The task TF1 is for the classification of fundus photographs as normal and abnormal, whereas TF2 is for DR grading. Similarly, the task TO1 classifies OCT-based images into four classes, whereas the task TO2 classifies images as normal and abnormal. The empirical results show that the model achieves competitive results for retinopathy classification and grading using multitask learning with WLSRL.","PeriodicalId":369792,"journal":{"name":"2023 4th International Conference on Computing, Mathematics and Engineering Technologies (iCoMET)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117067625","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 : 2023-03-17DOI: 10.1109/iCoMET57998.2023.10099204
Maneeba Ashraf, Muhammad Asif, M. Ahmad, Ahsan Ayaz, Ayesha Nasir, Umer Ahmad
Ransomware is a typical malware attack that has been increasing steadily over the last few years. It encrypts users' data or removes significant material. The attackers ask for money to unlock and decrypt the data. In this paper, an analysis of Ransomware attack and its detection techniques is presented. Initially, Ransomware detection techniques are classified based on their working principle. After that, a detailed comparative analysis is made to figure out the suitability of these techniques in different scenarios.
{"title":"Towards Classification and Analysis of Ransomware Detection Techniques","authors":"Maneeba Ashraf, Muhammad Asif, M. Ahmad, Ahsan Ayaz, Ayesha Nasir, Umer Ahmad","doi":"10.1109/iCoMET57998.2023.10099204","DOIUrl":"https://doi.org/10.1109/iCoMET57998.2023.10099204","url":null,"abstract":"Ransomware is a typical malware attack that has been increasing steadily over the last few years. It encrypts users' data or removes significant material. The attackers ask for money to unlock and decrypt the data. In this paper, an analysis of Ransomware attack and its detection techniques is presented. Initially, Ransomware detection techniques are classified based on their working principle. After that, a detailed comparative analysis is made to figure out the suitability of these techniques in different scenarios.","PeriodicalId":369792,"journal":{"name":"2023 4th International Conference on Computing, Mathematics and Engineering Technologies (iCoMET)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122634486","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：