Pub Date : 2021-09-13DOI: 10.1109/IICAIET51634.2021.9573997
Junqi Huang, T. Kumar, Haider A. F. Almurib
The paper proposes generic expressions for early design phase and accurate estimation of the performance of binary multipliers of $n$-bits in length using eight kinds of traditional adders. The performance of n-bits multipliers using different adders can be quickly assessed in theory by using proposed generic expressions without actual circuits. Performance parameters that are considered are namely the number of stages, gate counts, required area, energy dissipated and worst-case gate level delay. Full adder array is applied to design RCA (Ripple Carry Adder) based multiplier; the number of adder cells at different stages are found. Then, multi-length adder array is designed for multiplier using multi-bits adders; the number of adders with different lengths are analyzed at different stages. Meanwhile, proposed expressions are validated against actual designs; estimated results using proposed expressions show in good agreement with results of actual circuits. Finally, different multipliers are compared in terms of their performances by using proposed expressions. Multipliers using KSA (Kogge-Stone adder) and CSLA (Carry Select adder) require the highest area ($3370 mu m^{2}$ for $n=16$) and consume the highest energy dissipation (2.5E-13J). The RCA based multiplier requires the lowest number of gates, area ($1637.96 mu m^{2}$) and energy dissipation (1.2791E-13J). Also, the worst-case delay for KSA based multiplier and SA (Sklansky adder) based multiplier is lowest (only 60 gate level delays), while that for RCA based multiplier is highest (209 gate level delays).
本文提出了使用8种传统加法器对长度为$n$位的二进制乘法器进行早期设计和性能准确估计的通用表达式。使用不同加法器的n位乘法器的性能可以在理论上通过使用提出的通用表达式而无需实际电路来快速评估。考虑的性能参数是级数、门数、所需面积、能量消耗和最坏情况下的门电平延迟。采用全加法器阵列设计基于RCA (Ripple Carry adder)的乘法器;发现了不同阶段加法器细胞的数量。然后,采用多位加法器设计了多长度加法器阵列;分析了不同长度加法器在不同阶段的数量。同时,根据实际设计对所提出的表达式进行验证;用所提表达式估计的结果与实际电路的结果吻合较好。最后,使用建议的表达式比较了不同乘数器的性能。使用KSA (Kogge-Stone加法器)和CSLA(进位选择加法器)的乘法器需要最大的面积($3370 mu m^{2}$对于$n=16$)并消耗最高的能量消耗(2.5E-13J)。基于RCA的乘法器需要最少的门数、面积($1637.96 mu m^{2}$)和能量消耗(1.2791E-13J)。此外,基于KSA的乘法器和基于SA (Sklansky加法器)的乘法器的最坏情况延迟是最低的(只有60个门级延迟),而基于RCA的乘法器的最坏情况延迟是最高的(209个门级延迟)。
{"title":"Generic Expressions for Early Estimation of Performance of Binary Multipliers","authors":"Junqi Huang, T. Kumar, Haider A. F. Almurib","doi":"10.1109/IICAIET51634.2021.9573997","DOIUrl":"https://doi.org/10.1109/IICAIET51634.2021.9573997","url":null,"abstract":"The paper proposes generic expressions for early design phase and accurate estimation of the performance of binary multipliers of $n$-bits in length using eight kinds of traditional adders. The performance of n-bits multipliers using different adders can be quickly assessed in theory by using proposed generic expressions without actual circuits. Performance parameters that are considered are namely the number of stages, gate counts, required area, energy dissipated and worst-case gate level delay. Full adder array is applied to design RCA (Ripple Carry Adder) based multiplier; the number of adder cells at different stages are found. Then, multi-length adder array is designed for multiplier using multi-bits adders; the number of adders with different lengths are analyzed at different stages. Meanwhile, proposed expressions are validated against actual designs; estimated results using proposed expressions show in good agreement with results of actual circuits. Finally, different multipliers are compared in terms of their performances by using proposed expressions. Multipliers using KSA (Kogge-Stone adder) and CSLA (Carry Select adder) require the highest area ($3370 mu m^{2}$ for $n=16$) and consume the highest energy dissipation (2.5E-13J). The RCA based multiplier requires the lowest number of gates, area ($1637.96 mu m^{2}$) and energy dissipation (1.2791E-13J). Also, the worst-case delay for KSA based multiplier and SA (Sklansky adder) based multiplier is lowest (only 60 gate level delays), while that for RCA based multiplier is highest (209 gate level delays).","PeriodicalId":234229,"journal":{"name":"2021 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IICAIET)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116509967","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 : 2021-09-13DOI: 10.1109/IICAIET51634.2021.9573596
R. Sharan
Sleep disorders affect millions of people worldwide. Polysomnography (PSG) is a sleep study that is commonly used to diagnose sleep disorders, such as using sleep staging. However, PSG can be labor intensive, time consuming, expensive, and may not be easily available. Sleep and wake cycles can cause variation in heart rate and respiration which can be estimated using electrocardiogram (ECG), available as wearable sensors. As such, this work studies the use of single-lead ECG for detecting sleep and wake stages, in particular, using the heart rate variability (HRV) and ECG-derived respiration (EDR) signals. Various temporal and spectral descriptors are extracted from the HRV and EDR signals for this purpose. Sequential backward feature selection is employed to select the discriminative features for classification using logistic regression. The proposed method is evaluated on a dataset of more than 85 hours of ECG recordings from 16 subjects in leave-one-subject-out cross-validation. An accuracy of 75% ($text{AUC} =0.83$) is achieved using the EDR features in classifying sleep and wake stages. This increased to an accuracy of 80% ($text{AUC} =0.88$) when combined with HRV features. The proposed method demonstrates potential to be used for screening sleep disorders using ECG.
{"title":"ECG-Derived Respiration for Sleep-Wake Stage Classification","authors":"R. Sharan","doi":"10.1109/IICAIET51634.2021.9573596","DOIUrl":"https://doi.org/10.1109/IICAIET51634.2021.9573596","url":null,"abstract":"Sleep disorders affect millions of people worldwide. Polysomnography (PSG) is a sleep study that is commonly used to diagnose sleep disorders, such as using sleep staging. However, PSG can be labor intensive, time consuming, expensive, and may not be easily available. Sleep and wake cycles can cause variation in heart rate and respiration which can be estimated using electrocardiogram (ECG), available as wearable sensors. As such, this work studies the use of single-lead ECG for detecting sleep and wake stages, in particular, using the heart rate variability (HRV) and ECG-derived respiration (EDR) signals. Various temporal and spectral descriptors are extracted from the HRV and EDR signals for this purpose. Sequential backward feature selection is employed to select the discriminative features for classification using logistic regression. The proposed method is evaluated on a dataset of more than 85 hours of ECG recordings from 16 subjects in leave-one-subject-out cross-validation. An accuracy of 75% ($text{AUC} =0.83$) is achieved using the EDR features in classifying sleep and wake stages. This increased to an accuracy of 80% ($text{AUC} =0.88$) when combined with HRV features. The proposed method demonstrates potential to be used for screening sleep disorders using ECG.","PeriodicalId":234229,"journal":{"name":"2021 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IICAIET)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116821394","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 : 2021-09-13DOI: 10.1109/IICAIET51634.2021.9573534
A. Alsakati, C. Vaithilingam, K. Naidu, Gowthamraj Rajendran, J. Alnasseir, A. Jagadeeshwaran
System stability plays a significant role in the development of modern power systems. Power System Stabilizer (PSS) is an effective device often used to provide auxiliary damping to the oscillations by stabilizing the signals. Particle Swarm Optimization (PSO) is a popular and intelligent optimization technique used to solve various optimization problems. In this research work, the transient stability of the two-area four-machine system is improved using PSS1A. PSS1A is a single-input single-band power system stabilizer. The PSS1A parameters are optimized using PSO to enhance the stability and mitigate the oscillations. The simulation results of the relative power angle of Synchronous Generators (SGs) show that the transient stability is significantly improved with Optimized PSS1A (O-PSS1A), and the oscillations are mitigated. The maximum relative power angle of generator 1 and generator 2, with respect to generator 4, reduced by 29.5% and 33.6% respectively with the proposed O-PSS1A. Similarly, a reduction was obtained in settling time of both generator 1 and generator 2 at 4.94 s and 3.66 s compared to the system without PSS which was unstable.
{"title":"Particle Swarm Optimization for Tuning Power System Stabilizer towards Transient Stability Improvement in Power System Network","authors":"A. Alsakati, C. Vaithilingam, K. Naidu, Gowthamraj Rajendran, J. Alnasseir, A. Jagadeeshwaran","doi":"10.1109/IICAIET51634.2021.9573534","DOIUrl":"https://doi.org/10.1109/IICAIET51634.2021.9573534","url":null,"abstract":"System stability plays a significant role in the development of modern power systems. Power System Stabilizer (PSS) is an effective device often used to provide auxiliary damping to the oscillations by stabilizing the signals. Particle Swarm Optimization (PSO) is a popular and intelligent optimization technique used to solve various optimization problems. In this research work, the transient stability of the two-area four-machine system is improved using PSS1A. PSS1A is a single-input single-band power system stabilizer. The PSS1A parameters are optimized using PSO to enhance the stability and mitigate the oscillations. The simulation results of the relative power angle of Synchronous Generators (SGs) show that the transient stability is significantly improved with Optimized PSS1A (O-PSS1A), and the oscillations are mitigated. The maximum relative power angle of generator 1 and generator 2, with respect to generator 4, reduced by 29.5% and 33.6% respectively with the proposed O-PSS1A. Similarly, a reduction was obtained in settling time of both generator 1 and generator 2 at 4.94 s and 3.66 s compared to the system without PSS which was unstable.","PeriodicalId":234229,"journal":{"name":"2021 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IICAIET)","volume":"137 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114597639","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 : 2021-09-13DOI: 10.1109/IICAIET51634.2021.9573571
Noor Atikah Mat Abir, Nor Bakiah Abd Warif, Nurezayana Zainal
As society is ever more dependent on technology and digital media, the law depends more on digital forensics to find, keep, evaluate, and analyze digital evidence such as digital images and digital documents. The effective image editing application that constantly improves allows the user to change the image material or alter the image effortlessly. Copy-move forgery (CMF) is a very difficult form of forgery to detect. CMF involves copying part of an image and pasting it into one or more regions of the same image. However, the existing Copy-Move Forgery Detection (CMFD) method was only utilized on the existing image dataset, while social media images are on the common media today. In this paper, the PatchMatch-based CMFD method is evaluated with different platforms of social media images: Facebook, WhatsApp, and Twitter. The average performance generated by the PatchMatch-based CMFD method is 91% for the existing CMFD dataset. By replacing the dataset with the social media images dataset, the average performance slightly decreases to 84%.
{"title":"An Evaluation of Patch Match-based Copy-Move Forgery Detection (CMFD) on Social Media Images","authors":"Noor Atikah Mat Abir, Nor Bakiah Abd Warif, Nurezayana Zainal","doi":"10.1109/IICAIET51634.2021.9573571","DOIUrl":"https://doi.org/10.1109/IICAIET51634.2021.9573571","url":null,"abstract":"As society is ever more dependent on technology and digital media, the law depends more on digital forensics to find, keep, evaluate, and analyze digital evidence such as digital images and digital documents. The effective image editing application that constantly improves allows the user to change the image material or alter the image effortlessly. Copy-move forgery (CMF) is a very difficult form of forgery to detect. CMF involves copying part of an image and pasting it into one or more regions of the same image. However, the existing Copy-Move Forgery Detection (CMFD) method was only utilized on the existing image dataset, while social media images are on the common media today. In this paper, the PatchMatch-based CMFD method is evaluated with different platforms of social media images: Facebook, WhatsApp, and Twitter. The average performance generated by the PatchMatch-based CMFD method is 91% for the existing CMFD dataset. By replacing the dataset with the social media images dataset, the average performance slightly decreases to 84%.","PeriodicalId":234229,"journal":{"name":"2021 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IICAIET)","volume":"366 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134406181","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 : 2021-09-13DOI: 10.1109/IICAIET51634.2021.9573986
M. Yaakop, S. Yaacob, A. A. Jamil, S. A. Bakar, Mohd. Fauzi Abu Hassan, A. S. Pri
Drowsiness detection has received a great deal of attention, and there are numerous EEG-based techniques for it. The signals are initially filtered, conditioned, and features are abstracted in this method, which focuses on post-processing, to assess the driver's drowsiness status. This method is frequently used, especially when the procedure's output yields odd results, as indicated in the literature. When a subject is in a dynamic position, such as driving when movement cannot be prevented or minimized, EEG data is difficult to get, and EEG signals are prone to artifacts such as muscle and head movement, among other things. Filtering is a software method for removing physical artifacts throughout the pre-and post-processing stages. This technique will take some time to develop and will have an impact on the overall detection time of the system. Algorithms for logic determination are used to determine whether the EEG probe's logic activity is active or inactive and to interpret it as drowsy. Data was collected from five healthy people aged 20 to 27 to put this technique to the test. Participants were instructed to continue driving while EEG data were collected and compared to sensor probe output to determine which wavelength best reflected their weariness. Sensory Logic monitors brain activity by measuring the strength of electrons gathered in a particular cortical location. When the two detection procedures are compared, the PSD approach has higher sensitivity and accuracy for detecting drowsiness, but the Sensor Boolean output falls short in the detection spectrum, as proven later.
{"title":"Detection of Drowsiness using EEG Probes Sensory Logic Signals Activeness Topology","authors":"M. Yaakop, S. Yaacob, A. A. Jamil, S. A. Bakar, Mohd. Fauzi Abu Hassan, A. S. Pri","doi":"10.1109/IICAIET51634.2021.9573986","DOIUrl":"https://doi.org/10.1109/IICAIET51634.2021.9573986","url":null,"abstract":"Drowsiness detection has received a great deal of attention, and there are numerous EEG-based techniques for it. The signals are initially filtered, conditioned, and features are abstracted in this method, which focuses on post-processing, to assess the driver's drowsiness status. This method is frequently used, especially when the procedure's output yields odd results, as indicated in the literature. When a subject is in a dynamic position, such as driving when movement cannot be prevented or minimized, EEG data is difficult to get, and EEG signals are prone to artifacts such as muscle and head movement, among other things. Filtering is a software method for removing physical artifacts throughout the pre-and post-processing stages. This technique will take some time to develop and will have an impact on the overall detection time of the system. Algorithms for logic determination are used to determine whether the EEG probe's logic activity is active or inactive and to interpret it as drowsy. Data was collected from five healthy people aged 20 to 27 to put this technique to the test. Participants were instructed to continue driving while EEG data were collected and compared to sensor probe output to determine which wavelength best reflected their weariness. Sensory Logic monitors brain activity by measuring the strength of electrons gathered in a particular cortical location. When the two detection procedures are compared, the PSD approach has higher sensitivity and accuracy for detecting drowsiness, but the Sensor Boolean output falls short in the detection spectrum, as proven later.","PeriodicalId":234229,"journal":{"name":"2021 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IICAIET)","volume":"65 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134320316","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 : 2021-09-13DOI: 10.1109/IICAIET51634.2021.9573579
Mounita Ghosh, Ferdib-Al-Islam
Type 2 diabetes mellitus is a severe disease in which the pancreas' insulin does not act correctly. In the United Kingdom, type 2 diabetes affects around 90% of diabetics. It is a severe ailment that might last a lifetime. Type 2 diabetes has no known cure. However, with the proper diagnosis at an early stage, type 2 diabetes may be managed, and the chance of getting it is reduced. In this research, machine learning has been applied to detect the presence of type 2 diabetes in patients. Exploratory Data Analysis has been performed to uncover the insights of the type 2 diabetes prediction dataset. Several classification algorithms - Support Vector Machine, Random Forest, and XGBoost algorithm were applied, and then feature importance scores were also computed to understand the feature impact on the development of the machine learning model. XGBoost model achieved better execution in different metrics like accuracy (100%), precision (100%), and recall (100%) and outperformed previous works.
{"title":"NDM-Finder: A Machine Learning Based Approach for Type-2 (Neonatal) Diabetes Mellitus Prediction","authors":"Mounita Ghosh, Ferdib-Al-Islam","doi":"10.1109/IICAIET51634.2021.9573579","DOIUrl":"https://doi.org/10.1109/IICAIET51634.2021.9573579","url":null,"abstract":"Type 2 diabetes mellitus is a severe disease in which the pancreas' insulin does not act correctly. In the United Kingdom, type 2 diabetes affects around 90% of diabetics. It is a severe ailment that might last a lifetime. Type 2 diabetes has no known cure. However, with the proper diagnosis at an early stage, type 2 diabetes may be managed, and the chance of getting it is reduced. In this research, machine learning has been applied to detect the presence of type 2 diabetes in patients. Exploratory Data Analysis has been performed to uncover the insights of the type 2 diabetes prediction dataset. Several classification algorithms - Support Vector Machine, Random Forest, and XGBoost algorithm were applied, and then feature importance scores were also computed to understand the feature impact on the development of the machine learning model. XGBoost model achieved better execution in different metrics like accuracy (100%), precision (100%), and recall (100%) and outperformed previous works.","PeriodicalId":234229,"journal":{"name":"2021 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IICAIET)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129912225","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 : 2021-09-13DOI: 10.1109/IICAIET51634.2021.9573706
J. Tan, K. Lim, C. Lee
Herbs are plants with savory or aromatic properties that are widely used for flavoring, food, medicine or perfume. The worldwide use of herbal products for healthcare has increased tremendously over the past decades. The plethora of herb species makes recognizing the herbs remains a challenge. This has spurred great interests among the researchers on pursuing artificial intelligent methods for herb classification. This paper presents a convolutional neural network (CNN) for herb classification. The proposed CNN consists of two convolution layers, two max pooling layers, a fully-connected layer and a softmax layer. The ReLU activation function and dropout regularization are leveraged to improve the performance of the proposed CNN. A dataset with 4067 herb images was collected for the evaluation purposes. The proposed CNN model achieves an accuracy of above 93% despite the fact that some herbs are visually similar.
{"title":"Herb Classification with Convolutional Neural Network","authors":"J. Tan, K. Lim, C. Lee","doi":"10.1109/IICAIET51634.2021.9573706","DOIUrl":"https://doi.org/10.1109/IICAIET51634.2021.9573706","url":null,"abstract":"Herbs are plants with savory or aromatic properties that are widely used for flavoring, food, medicine or perfume. The worldwide use of herbal products for healthcare has increased tremendously over the past decades. The plethora of herb species makes recognizing the herbs remains a challenge. This has spurred great interests among the researchers on pursuing artificial intelligent methods for herb classification. This paper presents a convolutional neural network (CNN) for herb classification. The proposed CNN consists of two convolution layers, two max pooling layers, a fully-connected layer and a softmax layer. The ReLU activation function and dropout regularization are leveraged to improve the performance of the proposed CNN. A dataset with 4067 herb images was collected for the evaluation purposes. The proposed CNN model achieves an accuracy of above 93% despite the fact that some herbs are visually similar.","PeriodicalId":234229,"journal":{"name":"2021 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IICAIET)","volume":"31 6","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120903080","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 : 2021-09-13DOI: 10.1109/IICAIET51634.2021.9573886
M. K. Tan, Norman Lim, Nurul Izyan Kamaruddin, Kit Guan Lim, Soo Siang Yang, K. Teo
This paper proposes artificial neural network (ANN) based maximum power point tracking (MPPT) controller to maximize the energy harvested by a grid-connected photovoltaic (PV) system under various environmental conditions. Due to the non-linear characteristics, PV system will exhibit multiple peaks when the PV array receives non-uniform irradiance. As such, the conventional perturb and observe (P&O) MPPT controller will be trapped at local maximum power point (MPP). Therefore, this paper aims to integrate ANN into MPPT controller to improve the effectiveness of the MPPT controller in tracking the global MPP. The effectiveness of the proposed method is tested under uniform and non-uniform irradiance conditions, and the performances are compared with the conventional P&O. The simulation results show the proposed method able to track the global MPP even the PV system exhibits multiple peaks under non-uniform condition, whereas the conventional P&O is trapped at local MPP. Thus, the proposed algorithm is able to harvest much energy as compared to the conventional method.
{"title":"Optimization of Photovoltaic Energy Harvesting using Artificial Neural Network","authors":"M. K. Tan, Norman Lim, Nurul Izyan Kamaruddin, Kit Guan Lim, Soo Siang Yang, K. Teo","doi":"10.1109/IICAIET51634.2021.9573886","DOIUrl":"https://doi.org/10.1109/IICAIET51634.2021.9573886","url":null,"abstract":"This paper proposes artificial neural network (ANN) based maximum power point tracking (MPPT) controller to maximize the energy harvested by a grid-connected photovoltaic (PV) system under various environmental conditions. Due to the non-linear characteristics, PV system will exhibit multiple peaks when the PV array receives non-uniform irradiance. As such, the conventional perturb and observe (P&O) MPPT controller will be trapped at local maximum power point (MPP). Therefore, this paper aims to integrate ANN into MPPT controller to improve the effectiveness of the MPPT controller in tracking the global MPP. The effectiveness of the proposed method is tested under uniform and non-uniform irradiance conditions, and the performances are compared with the conventional P&O. The simulation results show the proposed method able to track the global MPP even the PV system exhibits multiple peaks under non-uniform condition, whereas the conventional P&O is trapped at local MPP. Thus, the proposed algorithm is able to harvest much energy as compared to the conventional method.","PeriodicalId":234229,"journal":{"name":"2021 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IICAIET)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116595409","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 : 2021-09-13DOI: 10.1109/IICAIET51634.2021.9573574
Md. Farhad Hossain, Md. Ariful Islam, Syed Naimatullah Hussain, Debprosad Das, Ruhul Amin, M. Alam
Brain tumor can cause the creation of most aggressive cancer, with a much shorter life expectancy in most advanced stages, unless identified and treated accordingly. In earlier, radiologists have to manually identify the tumors from MRI images or other imaging types. That is both time consuming and threatening to the misclassification that could affect the recovery plan of a patient. Technological innovations and machine learning assist radiologists to detect tumors without invasive procedures. One of the machine learning algorithms that has been shown to be effective at image segmentation and classification is the convolutional neural network (CNN). In this proposed work, a novel CNN architecture was used on a publicly available figshare dataset to identify three brain tumor types. The proposed CNN architecture outperformed most state-of-the-art approaches, achieving a classification accuracy of 96.90 %. Precision, recall, and F1-score are some of the other evaluation metrics used in the study. In addition, the paper includes an in-depth analysis of misclassifications.
{"title":"Brain Tumor Classification from MRI Images Using Convolutional Neural Network","authors":"Md. Farhad Hossain, Md. Ariful Islam, Syed Naimatullah Hussain, Debprosad Das, Ruhul Amin, M. Alam","doi":"10.1109/IICAIET51634.2021.9573574","DOIUrl":"https://doi.org/10.1109/IICAIET51634.2021.9573574","url":null,"abstract":"Brain tumor can cause the creation of most aggressive cancer, with a much shorter life expectancy in most advanced stages, unless identified and treated accordingly. In earlier, radiologists have to manually identify the tumors from MRI images or other imaging types. That is both time consuming and threatening to the misclassification that could affect the recovery plan of a patient. Technological innovations and machine learning assist radiologists to detect tumors without invasive procedures. One of the machine learning algorithms that has been shown to be effective at image segmentation and classification is the convolutional neural network (CNN). In this proposed work, a novel CNN architecture was used on a publicly available figshare dataset to identify three brain tumor types. The proposed CNN architecture outperformed most state-of-the-art approaches, achieving a classification accuracy of 96.90 %. Precision, recall, and F1-score are some of the other evaluation metrics used in the study. In addition, the paper includes an in-depth analysis of misclassifications.","PeriodicalId":234229,"journal":{"name":"2021 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IICAIET)","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124180050","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 : 2021-09-13DOI: 10.1109/IICAIET51634.2021.9573623
Kok Luong Goh, G. Ng, Muzaffar Hamzah, S. Chai
Semi-automated segmentation, also known as interactive image segmentation, is an algorithm that extracts a region of interest (ROI) from an image based on user input. The said algorithm will be fed the user input information repeatedly until the required region of interest is successfully segmented. Pre-processing steps can be used to speed up the segmentation process while improving the end result. The use of superpixels is one example of such pre-processing step. A superpixel is a group of pixels that share similar characteristics such as texture and colour. Despite the fact that it is used as a pre-processing step in many interactive segmentation algorithms, less studies had been conducted to assess the effects of the size of superpixels required by interactive segmentation algorithms to achieve an optimal result. Therefore, the purpose of this research is to address this issue in order to bridge this research gap. This study will be performed using the Maximum Similarity based region merging (MSRM) with input strokes on selected images from the Berkeleys and Grabcut image data sets, generated by superpixels extractions via energy-driven samples (SEEDS We infer from this research that an image with a minimum of 500 superpixels will aid the interactive segmentation algorithm in producing a decent segmentation result with pixel accuracy of 0.963, F-score of 0.844, and Jaccard index of 0.756. When the superpixels for an image are raised to 10,000, the segmentation results degrade. In conclusion, the size of the superpixels would have an impact on the final segmentation results.
{"title":"Sizes of Superpixels and their Effect on Interactive Segmentation","authors":"Kok Luong Goh, G. Ng, Muzaffar Hamzah, S. Chai","doi":"10.1109/IICAIET51634.2021.9573623","DOIUrl":"https://doi.org/10.1109/IICAIET51634.2021.9573623","url":null,"abstract":"Semi-automated segmentation, also known as interactive image segmentation, is an algorithm that extracts a region of interest (ROI) from an image based on user input. The said algorithm will be fed the user input information repeatedly until the required region of interest is successfully segmented. Pre-processing steps can be used to speed up the segmentation process while improving the end result. The use of superpixels is one example of such pre-processing step. A superpixel is a group of pixels that share similar characteristics such as texture and colour. Despite the fact that it is used as a pre-processing step in many interactive segmentation algorithms, less studies had been conducted to assess the effects of the size of superpixels required by interactive segmentation algorithms to achieve an optimal result. Therefore, the purpose of this research is to address this issue in order to bridge this research gap. This study will be performed using the Maximum Similarity based region merging (MSRM) with input strokes on selected images from the Berkeleys and Grabcut image data sets, generated by superpixels extractions via energy-driven samples (SEEDS We infer from this research that an image with a minimum of 500 superpixels will aid the interactive segmentation algorithm in producing a decent segmentation result with pixel accuracy of 0.963, F-score of 0.844, and Jaccard index of 0.756. When the superpixels for an image are raised to 10,000, the segmentation results degrade. In conclusion, the size of the superpixels would have an impact on the final segmentation results.","PeriodicalId":234229,"journal":{"name":"2021 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IICAIET)","volume":"40 4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127439916","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}