Pub Date : 2021-10-23DOI: 10.1109/NILES53778.2021.9600508
S. Kapoulea, C. Psychalinos, A. Elwakil
Topologies of power-law proportional-integral controllers, which offer minimization of the active component count are presented in this work. This is achieved thanks to the utilization of RC networks, which approximate driving-point impedances described by power-law functions. Additional important features of the presented schemes are their capability of achieving minimization of the spread of passive elements, as well as of implementing values of order greater than one.
{"title":"Reduced Active Element Power-Law Proportional-Integral Controller Designs","authors":"S. Kapoulea, C. Psychalinos, A. Elwakil","doi":"10.1109/NILES53778.2021.9600508","DOIUrl":"https://doi.org/10.1109/NILES53778.2021.9600508","url":null,"abstract":"Topologies of power-law proportional-integral controllers, which offer minimization of the active component count are presented in this work. This is achieved thanks to the utilization of RC networks, which approximate driving-point impedances described by power-law functions. Additional important features of the presented schemes are their capability of achieving minimization of the spread of passive elements, as well as of implementing values of order greater than one.","PeriodicalId":249153,"journal":{"name":"2021 3rd Novel Intelligent and Leading Emerging Sciences Conference (NILES)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121911170","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 : 2021-10-23DOI: 10.1109/NILES53778.2021.9600556
A. Gamal, Khaled Bedda, Nada Ashraf, Salma Ayman, M. Abdallah, M. Rushdi
For the sake of proper diagnosis and treatment, accurate brain tumour segmentation is required. Because manual brain tumour segmentation is a time-consuming, costly, and subjective task, effective automated approaches for this purpose are generally desired. However, because brain tumours vary greatly in terms of location, shape, and size, establishing automatic segmentation algorithms has remained challenging throughout the years. Automatic segmentation of brain tumour is the process of separating abnormal tissues from normal tissues, such as white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF). Brian segmentation needs typically to be carried out for different image modalities in order to reveal important metabolic and physiological information. These modalities include positron emission tomography (PET), computer tomography (CT) image and magnetic resonance image (MRI). Multimodal imaging techniques (such as PET/CT and PET/MRI) that combine the information from multiple imaging modalities contribute more for accurate brain tumour segmentation. In this work, we introduce a deep learning framework for automated segmentation of 3D brain tumors that can save physicians time and provide an accurate reproducible solution for further tumor analysis and monitoring. In particular, a 3D U-Net was trained on brain MRI data obtained from the 2018 Brain tumor Image Segmentation (BraTS) challenge. Three optimizers (RMSProp, Adam and Nadam) and three loss functions (Dice loss, focal Tversky loss, Log-Cosh loss functions) were used. We demonstrated that some loss functions and optimizers combinations perform better than other ones. For example, using the Log-Cosh loss function along with RMSProp optimizer resulted in the highest Dice coefficient, 0.75. Indeed, we also optimized the network hyperparameters in order to enhance the segmentation outcomes. These results demonstrate the feasibility and effectiveness of the proposed deep learning scheme with optimized hyperparemeters and appropriate selection of the optimizer and loss function.
{"title":"Brain Tumor Segmentation using 3D U-Net with Hyperparameter Optimization","authors":"A. Gamal, Khaled Bedda, Nada Ashraf, Salma Ayman, M. Abdallah, M. Rushdi","doi":"10.1109/NILES53778.2021.9600556","DOIUrl":"https://doi.org/10.1109/NILES53778.2021.9600556","url":null,"abstract":"For the sake of proper diagnosis and treatment, accurate brain tumour segmentation is required. Because manual brain tumour segmentation is a time-consuming, costly, and subjective task, effective automated approaches for this purpose are generally desired. However, because brain tumours vary greatly in terms of location, shape, and size, establishing automatic segmentation algorithms has remained challenging throughout the years. Automatic segmentation of brain tumour is the process of separating abnormal tissues from normal tissues, such as white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF). Brian segmentation needs typically to be carried out for different image modalities in order to reveal important metabolic and physiological information. These modalities include positron emission tomography (PET), computer tomography (CT) image and magnetic resonance image (MRI). Multimodal imaging techniques (such as PET/CT and PET/MRI) that combine the information from multiple imaging modalities contribute more for accurate brain tumour segmentation. In this work, we introduce a deep learning framework for automated segmentation of 3D brain tumors that can save physicians time and provide an accurate reproducible solution for further tumor analysis and monitoring. In particular, a 3D U-Net was trained on brain MRI data obtained from the 2018 Brain tumor Image Segmentation (BraTS) challenge. Three optimizers (RMSProp, Adam and Nadam) and three loss functions (Dice loss, focal Tversky loss, Log-Cosh loss functions) were used. We demonstrated that some loss functions and optimizers combinations perform better than other ones. For example, using the Log-Cosh loss function along with RMSProp optimizer resulted in the highest Dice coefficient, 0.75. Indeed, we also optimized the network hyperparameters in order to enhance the segmentation outcomes. These results demonstrate the feasibility and effectiveness of the proposed deep learning scheme with optimized hyperparemeters and appropriate selection of the optimizer and loss function.","PeriodicalId":249153,"journal":{"name":"2021 3rd Novel Intelligent and Leading Emerging Sciences Conference (NILES)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122203927","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 : 2021-10-23DOI: 10.1109/NILES53778.2021.9600512
Ali Amin, Salmeen Bahnasy, K. Elghamry, A. Samir, A. Emad, M. Darweesh, A. El-Sherif
Creating a model to detect freely moving fish underwater in real-time is a challenging process for two main reasons. First, the available datasets suffer from some limitations that severely affect the results of the detection models operating in challenging and blurry environments. These models should be able to capture all of the fish movement given different types of surroundings. Second, choosing the convenient detection model system which matches the desired requirements from having high accuracy with satisfying frames per second (FPS). To overcome the first challenge, a new dataset was created by extracting 1800 frames from videos and manually annotating them to overcome the different background issues and the complex movements and orientations of the fish. Regarding the second challenge and after comparing between different object detection systems, YOLOv3 was chosen as it proved to achieve high accuracy among other systems. The proposed approach scored 76.81% using (mean average precision) mAP as an accuracy metric and 89.17% using F-score, which is considered one of the most accurate outcomes among the literature. Moreover, the model rate is 12 FPS which is satisfying for real-time.
{"title":"Real-Time Fish Detection Approach on Self-Built Dataset Based on YOLOv3","authors":"Ali Amin, Salmeen Bahnasy, K. Elghamry, A. Samir, A. Emad, M. Darweesh, A. El-Sherif","doi":"10.1109/NILES53778.2021.9600512","DOIUrl":"https://doi.org/10.1109/NILES53778.2021.9600512","url":null,"abstract":"Creating a model to detect freely moving fish underwater in real-time is a challenging process for two main reasons. First, the available datasets suffer from some limitations that severely affect the results of the detection models operating in challenging and blurry environments. These models should be able to capture all of the fish movement given different types of surroundings. Second, choosing the convenient detection model system which matches the desired requirements from having high accuracy with satisfying frames per second (FPS). To overcome the first challenge, a new dataset was created by extracting 1800 frames from videos and manually annotating them to overcome the different background issues and the complex movements and orientations of the fish. Regarding the second challenge and after comparing between different object detection systems, YOLOv3 was chosen as it proved to achieve high accuracy among other systems. The proposed approach scored 76.81% using (mean average precision) mAP as an accuracy metric and 89.17% using F-score, which is considered one of the most accurate outcomes among the literature. Moreover, the model rate is 12 FPS which is satisfying for real-time.","PeriodicalId":249153,"journal":{"name":"2021 3rd Novel Intelligent and Leading Emerging Sciences Conference (NILES)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125900483","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 : 2021-10-23DOI: 10.1109/NILES53778.2021.9600513
Sara Ahmed, Nancy Alshaer, T. Ismail
Random numbers play an essential role in guaranteeing secrecy in most cryptographic systems. A chaotic optical signal is exploited to achieve high-speed random numbers. It could be generated by using one or more semiconductor lasers with external optical feedback. However, this system faces two major issues, high peak to average power ratio (PAPR) and parameter variations. These issues highly affected the randomness of the generated bitstreams. In this paper, we use a non-linear compression technique to compand the generated signal before it is quantized to avoid the effects of the PAPR. Also, we develop the post-processing stage by using advanced encryption standard (AES) algorithm feeds from two different generated bitstreams. These two integrated stages, non-linear quantization, and post-processing are configured to achieve a generation of a efficient random number guaranteed by NIST and DIEHARD statistical test suites. Finally, the proposed system is verified at parameter variation of ±20% tolerance including external mirror reflectivity, external cavity length, and normalized injection current. The results show that the proposed system could generate truly random numbers even with parameters configuration tolerance.
{"title":"Chaos-Based RNG using Semiconductor Lasers with Parameters Variation Tolerance","authors":"Sara Ahmed, Nancy Alshaer, T. Ismail","doi":"10.1109/NILES53778.2021.9600513","DOIUrl":"https://doi.org/10.1109/NILES53778.2021.9600513","url":null,"abstract":"Random numbers play an essential role in guaranteeing secrecy in most cryptographic systems. A chaotic optical signal is exploited to achieve high-speed random numbers. It could be generated by using one or more semiconductor lasers with external optical feedback. However, this system faces two major issues, high peak to average power ratio (PAPR) and parameter variations. These issues highly affected the randomness of the generated bitstreams. In this paper, we use a non-linear compression technique to compand the generated signal before it is quantized to avoid the effects of the PAPR. Also, we develop the post-processing stage by using advanced encryption standard (AES) algorithm feeds from two different generated bitstreams. These two integrated stages, non-linear quantization, and post-processing are configured to achieve a generation of a efficient random number guaranteed by NIST and DIEHARD statistical test suites. Finally, the proposed system is verified at parameter variation of ±20% tolerance including external mirror reflectivity, external cavity length, and normalized injection current. The results show that the proposed system could generate truly random numbers even with parameters configuration tolerance.","PeriodicalId":249153,"journal":{"name":"2021 3rd Novel Intelligent and Leading Emerging Sciences Conference (NILES)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125871336","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 : 2021-10-23DOI: 10.1109/NILES53778.2021.9600492
Soha Ahmed Ehssan Aly, Aya Hassanin, Saddam Bekhet
Sign languages is a critical requirement that helps deaf people to express their needs, feelings and emotions using a variety of hand gestures throughout their daily life. This language had evolved in parallel with spoken languages, however, it do not resemble its counterparts in the same way. Moreover, it is as complex as any other spoken language, as each sign language embodies hundreds of signs, that differs from the next by slight changes in hand shape, position, motion direction, face and body parts contributing to each sign. Unfortunately, sign languages are not globally standardized, where the language differs between countries and has its own vocabulary and varies although they might look similar. Furthermore, publicly available datasets are limited in quality and most of the available translation services are expensive, due to the required skilled human personnel. This paper proposes a deep learning approach for sign language detection that is finely tailored for the Egyptian sign language (special case of the generic sign language). The model is built to harnesses the power of convolutional and recurrent networks by integrating them together to better recognize the sign language spatio-temporal data-feed. In addition, the paper proposes the first Egyptian sign language dataset for emotion words and pronouns. The experimental results demonstrated the proposed approach promising results on the introduced dataset using combined CNN with RNN models.
{"title":"ESLDL: An Integrated Deep Learning Model for Egyptian Sign Language Recognition","authors":"Soha Ahmed Ehssan Aly, Aya Hassanin, Saddam Bekhet","doi":"10.1109/NILES53778.2021.9600492","DOIUrl":"https://doi.org/10.1109/NILES53778.2021.9600492","url":null,"abstract":"Sign languages is a critical requirement that helps deaf people to express their needs, feelings and emotions using a variety of hand gestures throughout their daily life. This language had evolved in parallel with spoken languages, however, it do not resemble its counterparts in the same way. Moreover, it is as complex as any other spoken language, as each sign language embodies hundreds of signs, that differs from the next by slight changes in hand shape, position, motion direction, face and body parts contributing to each sign. Unfortunately, sign languages are not globally standardized, where the language differs between countries and has its own vocabulary and varies although they might look similar. Furthermore, publicly available datasets are limited in quality and most of the available translation services are expensive, due to the required skilled human personnel. This paper proposes a deep learning approach for sign language detection that is finely tailored for the Egyptian sign language (special case of the generic sign language). The model is built to harnesses the power of convolutional and recurrent networks by integrating them together to better recognize the sign language spatio-temporal data-feed. In addition, the paper proposes the first Egyptian sign language dataset for emotion words and pronouns. The experimental results demonstrated the proposed approach promising results on the introduced dataset using combined CNN with RNN models.","PeriodicalId":249153,"journal":{"name":"2021 3rd Novel Intelligent and Leading Emerging Sciences Conference (NILES)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130576161","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 : 2021-10-23DOI: 10.1109/NILES53778.2021.9600518
Yusuf T. Elbadry, A. Guaily, M. Boraey, M. Abdelrahman
The active control of flow around an airfoil through morphing is numerically investigated. The lock-in phenomenon was predicted while using a fixed grid. Galerkin/Least-Squares Finite Element Method was used to simulate incompressible flow over an airfoil with leading edge morphing at a Reynolds number, $Re = 5000$, and angle of attack, $alpha = 6^{circ}$. The numerical simulation was carried out using the in-house FORTRAN code. The code was validated with the literature by simulating the flow over an oscillating cylinder. The paperwork implemented a locally oscillating surface on the airfoil with a deformation function. The non-dimensional oscillation frequency was varied in the range of [0.4 - 2.7] and the flow frequencies were analyzed. The primary and secondary frequencies were recorded at each simulation and the lock-in region is specified. The streamlines and vorticity contours are presented at two different excitation frequencies, specifically, $f_{e} = 1.0$ and $f_{e} = 2.5$. The streamlines and vorticity contours showed the formation of the vortices in both cases. The results show great accuracy for the Level-Set Method compared with the literature work that used the Arbitrary Lagrangian-Eulerian method, and the flow frequencies can be predicted accurately.
{"title":"Active Morphing Control of Airfoil At Low Reynolds Number Using Level-Set Method","authors":"Yusuf T. Elbadry, A. Guaily, M. Boraey, M. Abdelrahman","doi":"10.1109/NILES53778.2021.9600518","DOIUrl":"https://doi.org/10.1109/NILES53778.2021.9600518","url":null,"abstract":"The active control of flow around an airfoil through morphing is numerically investigated. The lock-in phenomenon was predicted while using a fixed grid. Galerkin/Least-Squares Finite Element Method was used to simulate incompressible flow over an airfoil with leading edge morphing at a Reynolds number, $Re = 5000$, and angle of attack, $alpha = 6^{circ}$. The numerical simulation was carried out using the in-house FORTRAN code. The code was validated with the literature by simulating the flow over an oscillating cylinder. The paperwork implemented a locally oscillating surface on the airfoil with a deformation function. The non-dimensional oscillation frequency was varied in the range of [0.4 - 2.7] and the flow frequencies were analyzed. The primary and secondary frequencies were recorded at each simulation and the lock-in region is specified. The streamlines and vorticity contours are presented at two different excitation frequencies, specifically, $f_{e} = 1.0$ and $f_{e} = 2.5$. The streamlines and vorticity contours showed the formation of the vortices in both cases. The results show great accuracy for the Level-Set Method compared with the literature work that used the Arbitrary Lagrangian-Eulerian method, and the flow frequencies can be predicted accurately.","PeriodicalId":249153,"journal":{"name":"2021 3rd Novel Intelligent and Leading Emerging Sciences Conference (NILES)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123286146","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 : 2021-10-23DOI: 10.1109/NILES53778.2021.9600524
Amr Medhat, M. Elattar, O. Fahmy
Interference management is one of the challenging tasks in Long-Term Evolution (LTE) technologies in Telecom Networks. One of these tasks is classifying interference problems affecting Uplink (UL) channel into different types. The interference classification problem can be formulated as an image classification task by converting the signal's power spectral density to an image. Convolutional Neural Networks (CNN) proved to have great success in image classification tasks. In this paper, different CNN architectures such as (VGG, MobileNet, RESNET) are used and assessed to classify the type of interference affecting the uplink channel in LTE. CNNs are characterized by their ability to detect and describe the abnormal behavior of UL channel which provided significant improvement over traditional rule-based systems. These rule-based systems rely on extracting domain driven features and classifying the interference using manually created rules by an expert. Our study shows that CNN yields 95% accuracy with training data. The end-to-end solution was deployed in Vodafone Group on Google Cloud Platform (GCP) to serve the different Local Markets.
{"title":"LTE Uplink Interference Inspection Using Convolutional Neural Networks","authors":"Amr Medhat, M. Elattar, O. Fahmy","doi":"10.1109/NILES53778.2021.9600524","DOIUrl":"https://doi.org/10.1109/NILES53778.2021.9600524","url":null,"abstract":"Interference management is one of the challenging tasks in Long-Term Evolution (LTE) technologies in Telecom Networks. One of these tasks is classifying interference problems affecting Uplink (UL) channel into different types. The interference classification problem can be formulated as an image classification task by converting the signal's power spectral density to an image. Convolutional Neural Networks (CNN) proved to have great success in image classification tasks. In this paper, different CNN architectures such as (VGG, MobileNet, RESNET) are used and assessed to classify the type of interference affecting the uplink channel in LTE. CNNs are characterized by their ability to detect and describe the abnormal behavior of UL channel which provided significant improvement over traditional rule-based systems. These rule-based systems rely on extracting domain driven features and classifying the interference using manually created rules by an expert. Our study shows that CNN yields 95% accuracy with training data. The end-to-end solution was deployed in Vodafone Group on Google Cloud Platform (GCP) to serve the different Local Markets.","PeriodicalId":249153,"journal":{"name":"2021 3rd Novel Intelligent and Leading Emerging Sciences Conference (NILES)","volume":"134 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123447745","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 : 2021-10-23DOI: 10.1109/NILES53778.2021.9600549
Youssef Abdelrahman, M. El-Salamony, Mohamed Khalifa
Most of the measurement devices in the university labs are not computerized. Hence, unsteady measurements are difficult to capture. In order to retrieve the measured data to computers, expensive data acquisition systems are needed to link these devices to computers. To overcome this issue a cost-efficient solution is proposed. This article proposes a methodology to convert the readings of LCDs of the various measuring devices into a digital form using computer vision. The procedure is successfully implemented and the results are presented.
{"title":"Visual Data Acquisition for Measuring Devices using Deep Learning","authors":"Youssef Abdelrahman, M. El-Salamony, Mohamed Khalifa","doi":"10.1109/NILES53778.2021.9600549","DOIUrl":"https://doi.org/10.1109/NILES53778.2021.9600549","url":null,"abstract":"Most of the measurement devices in the university labs are not computerized. Hence, unsteady measurements are difficult to capture. In order to retrieve the measured data to computers, expensive data acquisition systems are needed to link these devices to computers. To overcome this issue a cost-efficient solution is proposed. This article proposes a methodology to convert the readings of LCDs of the various measuring devices into a digital form using computer vision. The procedure is successfully implemented and the results are presented.","PeriodicalId":249153,"journal":{"name":"2021 3rd Novel Intelligent and Leading Emerging Sciences Conference (NILES)","volume":"122 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128028573","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 : 2021-10-23DOI: 10.1109/NILES53778.2021.9600522
B. A. Mohamed, Lamees N. Mahmoud, W. Al-Atabany, N. Salem
The precise detection and segmentation of cells in microscopic image sequences is an essential task in biomedical research, such as drug discovery and studying the development of tissues, organs, or entire organisms. However, the detection and segmentation of cells in phase contrast images with a halo and shade-off effects is still challenging. Lately, Mask Regional Convolutional Neural Network (Mask R-CNN) has been introduced for object detection and instance segmentation of natural images. This study investigates the efficacy of the Mask R-CNN to instantly detect and segment label-free microscopic images. The dataset used in this paper is taken from the ISBI cell tracking challenge. The Mask R-CNN is trained using different hyperparameters and compared to the U-Net model. Experimental results show that the Mask R-CNN model achieves 91.6 % when using ResNet-50 backbone and COCO weights.
{"title":"Instance Segmentation of 2D Label-Free Microscopic Images using Deep Learning","authors":"B. A. Mohamed, Lamees N. Mahmoud, W. Al-Atabany, N. Salem","doi":"10.1109/NILES53778.2021.9600522","DOIUrl":"https://doi.org/10.1109/NILES53778.2021.9600522","url":null,"abstract":"The precise detection and segmentation of cells in microscopic image sequences is an essential task in biomedical research, such as drug discovery and studying the development of tissues, organs, or entire organisms. However, the detection and segmentation of cells in phase contrast images with a halo and shade-off effects is still challenging. Lately, Mask Regional Convolutional Neural Network (Mask R-CNN) has been introduced for object detection and instance segmentation of natural images. This study investigates the efficacy of the Mask R-CNN to instantly detect and segment label-free microscopic images. The dataset used in this paper is taken from the ISBI cell tracking challenge. The Mask R-CNN is trained using different hyperparameters and compared to the U-Net model. Experimental results show that the Mask R-CNN model achieves 91.6 % when using ResNet-50 backbone and COCO weights.","PeriodicalId":249153,"journal":{"name":"2021 3rd Novel Intelligent and Leading Emerging Sciences Conference (NILES)","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131104203","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 : 2021-10-23DOI: 10.1109/NILES53778.2021.9600534
M. Salem, M. Elbanna, M. Abouelatta, Ahmed Saeed, A. Shaker
The simulation of quantum transport in DG-MOSFETs could be effectively accomplished by the Partial-Coupled Mode Space (PCMS) approach, which is realized by separating the odd and even modes solutions. This technique combines the merits of Coupled Mode Space (CMS) regarding the accuracy and Uncoupled Mode Space (UMS) as far as reducing computational time is concerned. In this work, a comparison study between PCMS using our developed FETMOSS simulator and CMS using Silvaco TCAD is carried out. The simulation is performed on a set of short-channel DG-MOSFETs. The accuracy at room temperature is found to be less than 8% along the whole range of the supply voltage. Based on this study, the PCMS approach in FETMOSS simulator is validated and proved to trace device performance in reasonable times compared to the TCAD high computational times.
{"title":"A Comparative Simulation Study of DG-MOSFETs: PCMS Approach in FETMOSS vs. CMS in Silvaco TCAD","authors":"M. Salem, M. Elbanna, M. Abouelatta, Ahmed Saeed, A. Shaker","doi":"10.1109/NILES53778.2021.9600534","DOIUrl":"https://doi.org/10.1109/NILES53778.2021.9600534","url":null,"abstract":"The simulation of quantum transport in DG-MOSFETs could be effectively accomplished by the Partial-Coupled Mode Space (PCMS) approach, which is realized by separating the odd and even modes solutions. This technique combines the merits of Coupled Mode Space (CMS) regarding the accuracy and Uncoupled Mode Space (UMS) as far as reducing computational time is concerned. In this work, a comparison study between PCMS using our developed FETMOSS simulator and CMS using Silvaco TCAD is carried out. The simulation is performed on a set of short-channel DG-MOSFETs. The accuracy at room temperature is found to be less than 8% along the whole range of the supply voltage. Based on this study, the PCMS approach in FETMOSS simulator is validated and proved to trace device performance in reasonable times compared to the TCAD high computational times.","PeriodicalId":249153,"journal":{"name":"2021 3rd Novel Intelligent and Leading Emerging Sciences Conference (NILES)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132487028","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：