Pub Date : 2022-12-19DOI: 10.1109/JAC-ECC56395.2022.10044073
O. Abdel-Rahim, Adel Bedair, D. Vinnikov, A. Chub, A. Blinov
A wide input voltage range isolated dc-dc isolated is utilized in this paper to demonstrate a parallel input series output partial power converter. The converter is operating well at wide range which is important feature for PV system, where the voltage is moving all over the year. The system principle of operation and simulation results are presented and discussed in the paper. The converter is well operating with more than 99% efficiency when partiality ratio between±15%. However, with high partiality like ± 30%, the converter is losing its main (a) features. An MPPT algorithm is moving the phase shift of the converter to extract the maximum power and design the partiality ratio.
{"title":"Partial Power Converter Based on Isolated Wide Input Range DC-DC Converter for Residential PV Applications","authors":"O. Abdel-Rahim, Adel Bedair, D. Vinnikov, A. Chub, A. Blinov","doi":"10.1109/JAC-ECC56395.2022.10044073","DOIUrl":"https://doi.org/10.1109/JAC-ECC56395.2022.10044073","url":null,"abstract":"A wide input voltage range isolated dc-dc isolated is utilized in this paper to demonstrate a parallel input series output partial power converter. The converter is operating well at wide range which is important feature for PV system, where the voltage is moving all over the year. The system principle of operation and simulation results are presented and discussed in the paper. The converter is well operating with more than 99% efficiency when partiality ratio between±15%. However, with high partiality like ± 30%, the converter is losing its main (a) features. An MPPT algorithm is moving the phase shift of the converter to extract the maximum power and design the partiality ratio.","PeriodicalId":326002,"journal":{"name":"2022 10th International Japan-Africa Conference on Electronics, Communications, and Computations (JAC-ECC)","volume":"96 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114735913","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 : 2022-12-19DOI: 10.1109/JAC-ECC56395.2022.10043965
Kibitok Abraham, M. Abdelwahab, M. Abo-Zahhad
Natural disasters continue to affect the world through wildfires, cyclones, earthquakes, and floods. The advent of photography has provided us with valuable images of how disasters happen and their impact. Many deep-learning models have been developed to classify images. However, the classification of natural disasters still lags. Through transfer learning, eleven existing deep learning models and two optimizers were adapted, analyzed and tested on images based on natural disasters. We explore the impact of data augmentation on deep learning model performance. Based on experimental results, ResNet-50 coupled with SGDM optimizer achieved an accuracy of 98.6%. However, AlexNet converge faster in 4109 seconds, compared to all adopted deep learning models.
{"title":"Image Classification of Natural Disasters Using Different Deep Learning Models","authors":"Kibitok Abraham, M. Abdelwahab, M. Abo-Zahhad","doi":"10.1109/JAC-ECC56395.2022.10043965","DOIUrl":"https://doi.org/10.1109/JAC-ECC56395.2022.10043965","url":null,"abstract":"Natural disasters continue to affect the world through wildfires, cyclones, earthquakes, and floods. The advent of photography has provided us with valuable images of how disasters happen and their impact. Many deep-learning models have been developed to classify images. However, the classification of natural disasters still lags. Through transfer learning, eleven existing deep learning models and two optimizers were adapted, analyzed and tested on images based on natural disasters. We explore the impact of data augmentation on deep learning model performance. Based on experimental results, ResNet-50 coupled with SGDM optimizer achieved an accuracy of 98.6%. However, AlexNet converge faster in 4109 seconds, compared to all adopted deep learning models.","PeriodicalId":326002,"journal":{"name":"2022 10th International Japan-Africa Conference on Electronics, Communications, and Computations (JAC-ECC)","volume":"253 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127362350","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 : 2022-12-19DOI: 10.1109/JAC-ECC56395.2022.10043968
Ziad Ibrahim, Yasmine Fahmy
Polar codes are part of the 5G technology enablers as they nearly achieve memoryless channels capacity. Recently, many researchers explored Machine-learning (ML) techniques to increase the effectiveness of polar codes decoders. Recurrent Neural Network Belief Propagation decoders (RNN-BP) demonstrated superior efficiency in fewer cycles than standard Polar decoders, as traditional Belief propagation decoders performed poorly throughout a limited number of iterations. In this paper, we present the first implementation of the RNN-BP on Field Programmable Gate Arrays (FPGA). Two implementations are presented in this paper. The first one uses a single processing unit, and the second one is enhanced by multiple processing units with a pipeline register. The multiple processing units’ design of RNN-BP shows much higher throughput which is 17x of belief propagation decoder implementation and 3x of the Soft-output CANcellation (SCAN) decoder implementation. We also achieved better BER than the two mentioned implementations 7. 08x than the original BP and 4. 5x better than the SCAN implementation. The combined memory and registers resource consumption in our design is less than the compared two implementations while consuming a larger number of look-up tables (LUTs).
{"title":"FPGA Implementation of Recurrent Neural Network-Based Polar Decoder","authors":"Ziad Ibrahim, Yasmine Fahmy","doi":"10.1109/JAC-ECC56395.2022.10043968","DOIUrl":"https://doi.org/10.1109/JAC-ECC56395.2022.10043968","url":null,"abstract":"Polar codes are part of the 5G technology enablers as they nearly achieve memoryless channels capacity. Recently, many researchers explored Machine-learning (ML) techniques to increase the effectiveness of polar codes decoders. Recurrent Neural Network Belief Propagation decoders (RNN-BP) demonstrated superior efficiency in fewer cycles than standard Polar decoders, as traditional Belief propagation decoders performed poorly throughout a limited number of iterations. In this paper, we present the first implementation of the RNN-BP on Field Programmable Gate Arrays (FPGA). Two implementations are presented in this paper. The first one uses a single processing unit, and the second one is enhanced by multiple processing units with a pipeline register. The multiple processing units’ design of RNN-BP shows much higher throughput which is 17x of belief propagation decoder implementation and 3x of the Soft-output CANcellation (SCAN) decoder implementation. We also achieved better BER than the two mentioned implementations 7. 08x than the original BP and 4. 5x better than the SCAN implementation. The combined memory and registers resource consumption in our design is less than the compared two implementations while consuming a larger number of look-up tables (LUTs).","PeriodicalId":326002,"journal":{"name":"2022 10th International Japan-Africa Conference on Electronics, Communications, and Computations (JAC-ECC)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126360769","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 : 2022-12-19DOI: 10.1109/JAC-ECC56395.2022.10044048
M. Fujii
A reference signal allocation scheme is described for dedicated energy beamforming in distributed wireless energy transfer. By allocating a common reference signal to multiple antennas as far apart as possible while preserving the randomness of the allocations, we aim to reduce the overhead of the frame structure, avoid the occurrence of long-lasting radio field strengths elsewhere, and reduce the computational complexity at the receiver. Our simulation results demonstrated that dedicated energy beamforming from an antenna-group subset was able to track the channel phase variations induced by a maximum velocity of 2.0 m/s and provided longer duration of high received power than per-antenna reference signal allocation.
{"title":"Dedicated Energy Beamforming in Distributed Antenna Systems Using Antenna-Group Common Reference Signals","authors":"M. Fujii","doi":"10.1109/JAC-ECC56395.2022.10044048","DOIUrl":"https://doi.org/10.1109/JAC-ECC56395.2022.10044048","url":null,"abstract":"A reference signal allocation scheme is described for dedicated energy beamforming in distributed wireless energy transfer. By allocating a common reference signal to multiple antennas as far apart as possible while preserving the randomness of the allocations, we aim to reduce the overhead of the frame structure, avoid the occurrence of long-lasting radio field strengths elsewhere, and reduce the computational complexity at the receiver. Our simulation results demonstrated that dedicated energy beamforming from an antenna-group subset was able to track the channel phase variations induced by a maximum velocity of 2.0 m/s and provided longer duration of high received power than per-antenna reference signal allocation.","PeriodicalId":326002,"journal":{"name":"2022 10th International Japan-Africa Conference on Electronics, Communications, and Computations (JAC-ECC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131333948","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 : 2022-12-19DOI: 10.1109/JAC-ECC56395.2022.10044040
Tokka Elsayed Ahmed, Walid A. Raslan, Atif Ghouniem, H. Abdel-Atty
Recently, the shipping industry and marine operations have been intensively reliable on the internet of things (IoT).IoT devices are used for the automation of shipping operations cargo tracking and security monitoring of the ships. In this paper, we focus on maintaining efficient security levels for marine communications. We propose new cryptographic methods based on the hyperchaotic sequence and the DNA sequence for image encryption. The proposed algorithm is applied to a smart buoy to enhance the security of the transmitted image through IoT systems in shipping industry monitoring. In addition, we evaluate and analyze the performance of three existing encryption algorithms for IoT (Blowfish, Advanced Encryption Standard (AES), and International Data Encryption Algorithm (IDEA)). Furthermore, a comparison was conducted between the three existing algorithms and the proposed one in terms of encryption time, correlation coefficient, and entropy before and after encryption. The simulation experiments show that the proposed cryptographic method outperforms the three existing algorithms in terms of quality, security, and robustness against noise as well as overall performance.
{"title":"Proposed HCDNA for Enhancing the Security of Marine Environment IoT Systems","authors":"Tokka Elsayed Ahmed, Walid A. Raslan, Atif Ghouniem, H. Abdel-Atty","doi":"10.1109/JAC-ECC56395.2022.10044040","DOIUrl":"https://doi.org/10.1109/JAC-ECC56395.2022.10044040","url":null,"abstract":"Recently, the shipping industry and marine operations have been intensively reliable on the internet of things (IoT).IoT devices are used for the automation of shipping operations cargo tracking and security monitoring of the ships. In this paper, we focus on maintaining efficient security levels for marine communications. We propose new cryptographic methods based on the hyperchaotic sequence and the DNA sequence for image encryption. The proposed algorithm is applied to a smart buoy to enhance the security of the transmitted image through IoT systems in shipping industry monitoring. In addition, we evaluate and analyze the performance of three existing encryption algorithms for IoT (Blowfish, Advanced Encryption Standard (AES), and International Data Encryption Algorithm (IDEA)). Furthermore, a comparison was conducted between the three existing algorithms and the proposed one in terms of encryption time, correlation coefficient, and entropy before and after encryption. The simulation experiments show that the proposed cryptographic method outperforms the three existing algorithms in terms of quality, security, and robustness against noise as well as overall performance.","PeriodicalId":326002,"journal":{"name":"2022 10th International Japan-Africa Conference on Electronics, Communications, and Computations (JAC-ECC)","volume":"81 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134447785","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 : 2022-12-19DOI: 10.1109/JAC-ECC56395.2022.10043892
Hazem M. Abd ElRaouf, Alhaytham M. Aref, A. Elsherif, Mohamed E. Khalifa
In this article, a mathematical model of the interaction between the immune system and a tumor cell employing an oncolytic viral therapy with time delay is discussed where the immune system has a dual function in the fight against cancer cells. Tumor cells that are sensitive to this sort of infection may become infected by viral infections that are developed to kill cancer cells but not healthy cells. Following oncolysis, the infected cancer cells release fresh viral infection particles to further destroy neighboring cancer cells. The boundedness of solutions of the model is derived. The time-delay differential equation theory is used to check whether the model’s equilibrium points are stable, moreover, the conditions that lead to a Hopf bifurcation are established. The results of the theoretical analysis are supported by numerical simulation.
{"title":"Stability and Hopf Bifurcation Analysis of a Tumor Immune Model of virus infection with Time-delay","authors":"Hazem M. Abd ElRaouf, Alhaytham M. Aref, A. Elsherif, Mohamed E. Khalifa","doi":"10.1109/JAC-ECC56395.2022.10043892","DOIUrl":"https://doi.org/10.1109/JAC-ECC56395.2022.10043892","url":null,"abstract":"In this article, a mathematical model of the interaction between the immune system and a tumor cell employing an oncolytic viral therapy with time delay is discussed where the immune system has a dual function in the fight against cancer cells. Tumor cells that are sensitive to this sort of infection may become infected by viral infections that are developed to kill cancer cells but not healthy cells. Following oncolysis, the infected cancer cells release fresh viral infection particles to further destroy neighboring cancer cells. The boundedness of solutions of the model is derived. The time-delay differential equation theory is used to check whether the model’s equilibrium points are stable, moreover, the conditions that lead to a Hopf bifurcation are established. The results of the theoretical analysis are supported by numerical simulation.","PeriodicalId":326002,"journal":{"name":"2022 10th International Japan-Africa Conference on Electronics, Communications, and Computations (JAC-ECC)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133623281","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 : 2022-12-19DOI: 10.1109/JAC-ECC56395.2022.10043970
M. Abo-Zahhad, Ashraf Mohamed Ali Hassan
Deep learning (DL) techniques were applied to ElectroCardioGrams (ECGs) for identifying abnormalities in heart diseases. The application of a convolutional neural network is discussed in this paper to perform the detection of coronary heart diseases by extracting QRS complexes and the error signal features. AlexNet, VGG19, ResNet50, GoogleNet, and NasNetLarge pretrained transfer learning models are adopted using the adaptive moment estimate (Adam) and stochastic gradient descent with momentum (SGDM) optimizers for training, testing, and validation. Here, 3 classes of heart diseases have been considered; namely arrhythmia, cardiomyopathy, and ischemia. The suggested method is aimed at automatically detecting these diseases based on ECG signals collected from MIT-BIH databases. The obtained results show that the Adam optimizer outperforms the SGDM optimizer for the five DL architectures. For AlexNet adopting Adam, the accuracy of detecting Arrhythmias, Ischaemia, and Cardiomyopathy is 98.2%, 95.9%, and 93.5% respectively. ResNet-50 and NasNetLarge with the same optimizer, have 98.0%, 96. 9%, 92.3 %, 97.9%, 95.7%, and 94.0 % accuracy in detecting Arrhythmias, Ischaemia, and Cardiomyopathy respectively. In addition, the subtraction of the QRS complexes from the clean ECG signal computationally outperforms the cutting-edge method based on using the continuous wavelet transform method. The reason is that the wavelet method is computationally expensive compared to the proposed QRS-complex subtraction method that results in integer error samples. So, the average execution time is significantly less.
{"title":"Classification of ECG Signals for Detecting Coronary Heart Diseases Using Deep Transfer Learning Techniques","authors":"M. Abo-Zahhad, Ashraf Mohamed Ali Hassan","doi":"10.1109/JAC-ECC56395.2022.10043970","DOIUrl":"https://doi.org/10.1109/JAC-ECC56395.2022.10043970","url":null,"abstract":"Deep learning (DL) techniques were applied to ElectroCardioGrams (ECGs) for identifying abnormalities in heart diseases. The application of a convolutional neural network is discussed in this paper to perform the detection of coronary heart diseases by extracting QRS complexes and the error signal features. AlexNet, VGG19, ResNet50, GoogleNet, and NasNetLarge pretrained transfer learning models are adopted using the adaptive moment estimate (Adam) and stochastic gradient descent with momentum (SGDM) optimizers for training, testing, and validation. Here, 3 classes of heart diseases have been considered; namely arrhythmia, cardiomyopathy, and ischemia. The suggested method is aimed at automatically detecting these diseases based on ECG signals collected from MIT-BIH databases. The obtained results show that the Adam optimizer outperforms the SGDM optimizer for the five DL architectures. For AlexNet adopting Adam, the accuracy of detecting Arrhythmias, Ischaemia, and Cardiomyopathy is 98.2%, 95.9%, and 93.5% respectively. ResNet-50 and NasNetLarge with the same optimizer, have 98.0%, 96. 9%, 92.3 %, 97.9%, 95.7%, and 94.0 % accuracy in detecting Arrhythmias, Ischaemia, and Cardiomyopathy respectively. In addition, the subtraction of the QRS complexes from the clean ECG signal computationally outperforms the cutting-edge method based on using the continuous wavelet transform method. The reason is that the wavelet method is computationally expensive compared to the proposed QRS-complex subtraction method that results in integer error samples. So, the average execution time is significantly less.","PeriodicalId":326002,"journal":{"name":"2022 10th International Japan-Africa Conference on Electronics, Communications, and Computations (JAC-ECC)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133165033","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 : 2022-12-19DOI: 10.1109/JAC-ECC56395.2022.10043887
S. Hashima, Kohei Hatano, E. M. Mohamed
Recent wireless communication systems, such as device-to-device (D2D) communications and internet of things (IoT), etc., support hybrid band frequencies to sustain the user demands in B5G/6G systems throughout switching between bands and avoid connection loss. This paper compares two online learning solutions for optimal band/channel assignment in hybrid radio frequency (WiFi and WiGig) and visible light communication (RF/VLC) wireless systems. In such scenarios, the multi-band source/transmitter (S/Tx) has no prior knowledge about distinct channel characteristics, including their transmission rates and consumed energy. Therefore, to extend its limited battery, the S/Tx has to target the best arm/band with the least possible consumed power. Hence, we compare two Multi Armed Bandit (MAB)-based solutions, which are costsubsidy MABs (CSMABs), where the S/Tx sacrifices with the highest reward in order to select the lowest cost arm/operating frequency and energy-aware MABs (EAMABs) where the cost term is amended only to the exploration term. In both methods, the S/Tx targets to maximize his cumulative payoff (transmission rate) and minimize his cost (battery expenditure due to the operating band/frequency). Numerical simulations indicate that proposed CS-MAB schemes outperform purely explored MABs via Thompson sampling (TS), upper Confidence bound (UCB), and benchmark multi-band election (MBE) approaches, correspondingly in terms of transmission rates and energy efficiency.
{"title":"Comparison of Energy Aware and Cost Subsidy Multi Armed Bandit Solutions for Supreme Channel Election in Hybrid Band Systems","authors":"S. Hashima, Kohei Hatano, E. M. Mohamed","doi":"10.1109/JAC-ECC56395.2022.10043887","DOIUrl":"https://doi.org/10.1109/JAC-ECC56395.2022.10043887","url":null,"abstract":"Recent wireless communication systems, such as device-to-device (D2D) communications and internet of things (IoT), etc., support hybrid band frequencies to sustain the user demands in B5G/6G systems throughout switching between bands and avoid connection loss. This paper compares two online learning solutions for optimal band/channel assignment in hybrid radio frequency (WiFi and WiGig) and visible light communication (RF/VLC) wireless systems. In such scenarios, the multi-band source/transmitter (S/Tx) has no prior knowledge about distinct channel characteristics, including their transmission rates and consumed energy. Therefore, to extend its limited battery, the S/Tx has to target the best arm/band with the least possible consumed power. Hence, we compare two Multi Armed Bandit (MAB)-based solutions, which are costsubsidy MABs (CSMABs), where the S/Tx sacrifices with the highest reward in order to select the lowest cost arm/operating frequency and energy-aware MABs (EAMABs) where the cost term is amended only to the exploration term. In both methods, the S/Tx targets to maximize his cumulative payoff (transmission rate) and minimize his cost (battery expenditure due to the operating band/frequency). Numerical simulations indicate that proposed CS-MAB schemes outperform purely explored MABs via Thompson sampling (TS), upper Confidence bound (UCB), and benchmark multi-band election (MBE) approaches, correspondingly in terms of transmission rates and energy efficiency.","PeriodicalId":326002,"journal":{"name":"2022 10th International Japan-Africa Conference on Electronics, Communications, and Computations (JAC-ECC)","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129465827","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 : 2022-12-19DOI: 10.1109/JAC-ECC56395.2022.10044005
Mohamed Osman, Mohamed Kamel, F. Mohammed, Tarek Hussein
The largest, most visible, and most exposed organ of the human body is the skin. Skin diseases impact about a third of the global population. Furthermore, many serious skin diseases, such as melanoma, may remain misdiagnosed for years since only experienced dermatologists can reliably diagnose them. As a result, some regions and demographic groups may be more negatively impacted than others in terms of access to such medical professionals. Given the range and possible hazards of skin diseases, democratizing access to accurate identification is critical. Fortunately, automated deep-learning systems have made significant advances in picture classification in recent years. In this paper we present SkinFormer, a vision transformer trained using strong augmentations and optimization techniques to ensure robustness and generalization. The dataset we use is a combination of 3 different datasets from previous works, and contains a total of 48,322 images. The choice of datasets and augmentations ensures strong generalization even on consumergrade hardware. We achieve a top-1 accuracy of 84.43% and a top-5 accuracy of 93.89%. We publish pre-trained model weights under an open source license.
{"title":"SkinFormer: Robust Vision Transformer for Automatic Skin Disease Identification","authors":"Mohamed Osman, Mohamed Kamel, F. Mohammed, Tarek Hussein","doi":"10.1109/JAC-ECC56395.2022.10044005","DOIUrl":"https://doi.org/10.1109/JAC-ECC56395.2022.10044005","url":null,"abstract":"The largest, most visible, and most exposed organ of the human body is the skin. Skin diseases impact about a third of the global population. Furthermore, many serious skin diseases, such as melanoma, may remain misdiagnosed for years since only experienced dermatologists can reliably diagnose them. As a result, some regions and demographic groups may be more negatively impacted than others in terms of access to such medical professionals. Given the range and possible hazards of skin diseases, democratizing access to accurate identification is critical. Fortunately, automated deep-learning systems have made significant advances in picture classification in recent years. In this paper we present SkinFormer, a vision transformer trained using strong augmentations and optimization techniques to ensure robustness and generalization. The dataset we use is a combination of 3 different datasets from previous works, and contains a total of 48,322 images. The choice of datasets and augmentations ensures strong generalization even on consumergrade hardware. We achieve a top-1 accuracy of 84.43% and a top-5 accuracy of 93.89%. We publish pre-trained model weights under an open source license.","PeriodicalId":326002,"journal":{"name":"2022 10th International Japan-Africa Conference on Electronics, Communications, and Computations (JAC-ECC)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127878370","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 : 2022-12-19DOI: 10.1109/JAC-ECC56395.2022.10043953
Rana Nader, Islam Nashaat, M. Dessouky
Analog and mixed-signal (AMS) layout implementation steps become more and more crucial in the IC design process. Hence, the layout design passes by several manual and time-consuming reviews to ensure that it meets the circuit designer’s expectations and that the circuit will work properly after fabrication. Automating the review process will guarantee the detection of any missing constraints assigned by the designer on the circuit and will reduce time and effort for the layout reviewer. We present an electronic design automation (EDA) tool that automatically checks matching patterns, metal widths to prevent electromigration and sensitive signal shielding.
{"title":"Automatic Analog and Mixed-Signal Physical Design Review","authors":"Rana Nader, Islam Nashaat, M. Dessouky","doi":"10.1109/JAC-ECC56395.2022.10043953","DOIUrl":"https://doi.org/10.1109/JAC-ECC56395.2022.10043953","url":null,"abstract":"Analog and mixed-signal (AMS) layout implementation steps become more and more crucial in the IC design process. Hence, the layout design passes by several manual and time-consuming reviews to ensure that it meets the circuit designer’s expectations and that the circuit will work properly after fabrication. Automating the review process will guarantee the detection of any missing constraints assigned by the designer on the circuit and will reduce time and effort for the layout reviewer. We present an electronic design automation (EDA) tool that automatically checks matching patterns, metal widths to prevent electromigration and sensitive signal shielding.","PeriodicalId":326002,"journal":{"name":"2022 10th International Japan-Africa Conference on Electronics, Communications, and Computations (JAC-ECC)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121043680","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：