Being a major agricultural country, a considerable amount of development depends on the agriculture of Bangladesh. As agriculture stays one of the main areas of the Bangladeshi economy, Bangladesh is attempting to become independent in producing food by creating successful developing agronomy. At the same time, plant leaf disease is quite natural and sometimes uncontrollable that causes damage of crops, as well as causing significant damage in the agronomy of Bangladesh. To prevent the problem, this work aims to classify several plant leaf diseases, specifically corn, grape, mango, and pepper, to diagnose the leaf diseases for proper early action to cure. We have also been able to classify by means of disease classification as a multi-class classification of those four plant leaves. Therefore, We have used Convolutional Neural Network (CNN) based Deep Learning models to analyze the results, and we have compared the scores of four CNN models: VGG-16, VGG-19, GoogLeNet, and our proposed model. Finally, our proposed model imparted better computation and achieved 99.91% accuracy. Furthermore, we have found that deep learning could be an appropriate approach to classify ill leaves of the plants from the healthy.
{"title":"Efficient Computation of Leaf Disease Classification Techniques using Deep Learning","authors":"Saifa Azmiri Mohona, Sakifa Aktar, Md. Martuza Ahamad","doi":"10.1109/ICEEE54059.2021.9718941","DOIUrl":"https://doi.org/10.1109/ICEEE54059.2021.9718941","url":null,"abstract":"Being a major agricultural country, a considerable amount of development depends on the agriculture of Bangladesh. As agriculture stays one of the main areas of the Bangladeshi economy, Bangladesh is attempting to become independent in producing food by creating successful developing agronomy. At the same time, plant leaf disease is quite natural and sometimes uncontrollable that causes damage of crops, as well as causing significant damage in the agronomy of Bangladesh. To prevent the problem, this work aims to classify several plant leaf diseases, specifically corn, grape, mango, and pepper, to diagnose the leaf diseases for proper early action to cure. We have also been able to classify by means of disease classification as a multi-class classification of those four plant leaves. Therefore, We have used Convolutional Neural Network (CNN) based Deep Learning models to analyze the results, and we have compared the scores of four CNN models: VGG-16, VGG-19, GoogLeNet, and our proposed model. Finally, our proposed model imparted better computation and achieved 99.91% accuracy. Furthermore, we have found that deep learning could be an appropriate approach to classify ill leaves of the plants from the healthy.","PeriodicalId":188366,"journal":{"name":"2021 3rd International Conference on Electrical & Electronic Engineering (ICEEE)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128788490","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 : 2021-12-22DOI: 10.1109/ICEEE54059.2021.9718786
M. Hasan, M. A. Kashem, Md. Jakirul Islam, Md. Zakir Hossain
Many real-world combinatorial optimization problems (COPs) are NP-hard and challenging to find the optimal solution using classical linear and convex optimization methods. In addition, the computational complexity of these optimization tasks increases exponentially with the increasing number of decision variables. A further difficulty can be also caused by the search space being intrinsically multimodal and non-convex. In such a case, an effective optimization method is required that can cope better with these problem characteristics. Genetic algorithm (GA) is a widely used method for COPs. The original GA and its variants have been used to solve a number of classic discrete optimization problems. Literature shows that the static mutation probability is commonly used for the GA and its variants which cause the imbalance between exploration and exploitation, limiting the performance of GA. To overcome this problem, this research proposes a time-varying mutation operator for GA. In this paper, the balance between exploration and exploitation of the proposed GA has been verified using the benchmark instances of a well-known combinatorial optimization problem i.e., the 0–1 knapsack problem. The numerical results show that the proposed GA can obtain better results with on average a significant number of function evaluations compared to the well-known metaheuristic methods.
{"title":"A Time-varying Mutation Operator for Balancing the Exploration and Exploitation Behaviours of Genetic Algorithm","authors":"M. Hasan, M. A. Kashem, Md. Jakirul Islam, Md. Zakir Hossain","doi":"10.1109/ICEEE54059.2021.9718786","DOIUrl":"https://doi.org/10.1109/ICEEE54059.2021.9718786","url":null,"abstract":"Many real-world combinatorial optimization problems (COPs) are NP-hard and challenging to find the optimal solution using classical linear and convex optimization methods. In addition, the computational complexity of these optimization tasks increases exponentially with the increasing number of decision variables. A further difficulty can be also caused by the search space being intrinsically multimodal and non-convex. In such a case, an effective optimization method is required that can cope better with these problem characteristics. Genetic algorithm (GA) is a widely used method for COPs. The original GA and its variants have been used to solve a number of classic discrete optimization problems. Literature shows that the static mutation probability is commonly used for the GA and its variants which cause the imbalance between exploration and exploitation, limiting the performance of GA. To overcome this problem, this research proposes a time-varying mutation operator for GA. In this paper, the balance between exploration and exploitation of the proposed GA has been verified using the benchmark instances of a well-known combinatorial optimization problem i.e., the 0–1 knapsack problem. The numerical results show that the proposed GA can obtain better results with on average a significant number of function evaluations compared to the well-known metaheuristic methods.","PeriodicalId":188366,"journal":{"name":"2021 3rd International Conference on Electrical & Electronic Engineering (ICEEE)","volume":"128 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128215716","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 : 2021-12-22DOI: 10.1109/ICEEE54059.2021.9718783
Nahin Ul Sadad, Afsana Afrin, Md. Nazrul Islam Mondal
Multiplication is one of the most common operations used in any program. Program working on massively large data always requires high computation power. In the age of big data, conventional general-purpose CPU based on Von Neumann architecture is no longer enough to satisfy high computation demand. Field Programmable Gate Array (FPGA) can perform hardware acceleration of any program. Since multiplier is the slowest component in any hardware accelerator, thus faster and re-configurable multiplier which can handle integers of any size must be implemented on FPGA. In this paper, we implemented both synchronous and asynchronous radix-2 booth multiplier using Verilog HDL on a Xilinx FPGA. We found that simulation time of asynchronous radix-2 booth multiplier is faster than synchronous radix-2 booth multiplier but synchronous radix-2 booth multiplier consumes fewer resources than asynchronous radix-2 booth multiplier.
{"title":"Synchronous and Asynchronous Implementation of Radix-2 Booth Multiplication Algorithm","authors":"Nahin Ul Sadad, Afsana Afrin, Md. Nazrul Islam Mondal","doi":"10.1109/ICEEE54059.2021.9718783","DOIUrl":"https://doi.org/10.1109/ICEEE54059.2021.9718783","url":null,"abstract":"Multiplication is one of the most common operations used in any program. Program working on massively large data always requires high computation power. In the age of big data, conventional general-purpose CPU based on Von Neumann architecture is no longer enough to satisfy high computation demand. Field Programmable Gate Array (FPGA) can perform hardware acceleration of any program. Since multiplier is the slowest component in any hardware accelerator, thus faster and re-configurable multiplier which can handle integers of any size must be implemented on FPGA. In this paper, we implemented both synchronous and asynchronous radix-2 booth multiplier using Verilog HDL on a Xilinx FPGA. We found that simulation time of asynchronous radix-2 booth multiplier is faster than synchronous radix-2 booth multiplier but synchronous radix-2 booth multiplier consumes fewer resources than asynchronous radix-2 booth multiplier.","PeriodicalId":188366,"journal":{"name":"2021 3rd International Conference on Electrical & Electronic Engineering (ICEEE)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133695524","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 : 2021-12-22DOI: 10.1109/ICEEE54059.2021.9718787
Shihab Ahammed, Kazi Sazzad Hossen, Ashraful Hossain Howlader
Of late, stanene and germanene having the effect of spin orbital coupling are characterized as a superconductive material at room temperature. These materials have been synthesized and investigated their low thermal conductivity in recent experimental studies. With the purpose of achieving diverse thermal properties, we have modeled and offered germanene/stanene heterobilayer. We have also characterized its in-plane thermal conduction with varying length. For the assessment its thermal properties, we employed a simulation method named reverse non equilibrium molecular dynamics. The nanosheet size in the x direction ranges from 20 to 300 nanometer. The amount of thermal transport of this heterobilayer is predicted to be 19.95 W m−1 K−1 over an unlimited length. In this work, the van der Waals thickness is used to predict this thermal transmission. The length of the nanosheet appears to boost the in-plane heat conduction of the germanene/stanene bilayer. For a better understanding of in-plane thermal conduction, the phonon density of states is determined. The characterization of germanene/stanene nanostructure proposed in this study would give a decent knowledge to make it a promising bilayer for the thermoelectric applications owing to its low thermal conductivity.
近年来，具有自旋轨道耦合效应的硅烯和锗烯在室温下被表征为超导材料。这些材料已被合成，并在最近的实验研究中研究了它们的低导热性。为了获得不同的热性能，我们模拟并提供了锗烯/stanene异质层。我们还描述了它的面内热传导随长度的变化。为了评估其热性能，我们采用了一种称为反向非平衡分子动力学的模拟方法。x方向的纳米片尺寸在20到300纳米之间。该异质层的热输运量预测为19.95 W m−1 K−1，长度不限。在这项工作中，范德华厚度被用来预测这种热传递。纳米片的长度似乎促进了锗烯/烯双分子层的平面内热传导。为了更好地理解面内热传导，确定了态声子密度。本研究提出的锗烯/stanene纳米结构的表征将使其具有良好的知识，使其成为热电应用的有前途的双层材料，因为它的低导热性。
{"title":"Length dependent thermal conduction in germanene/stanene heterobilayer by using molecular dynamics simulations","authors":"Shihab Ahammed, Kazi Sazzad Hossen, Ashraful Hossain Howlader","doi":"10.1109/ICEEE54059.2021.9718787","DOIUrl":"https://doi.org/10.1109/ICEEE54059.2021.9718787","url":null,"abstract":"Of late, stanene and germanene having the effect of spin orbital coupling are characterized as a superconductive material at room temperature. These materials have been synthesized and investigated their low thermal conductivity in recent experimental studies. With the purpose of achieving diverse thermal properties, we have modeled and offered germanene/stanene heterobilayer. We have also characterized its in-plane thermal conduction with varying length. For the assessment its thermal properties, we employed a simulation method named reverse non equilibrium molecular dynamics. The nanosheet size in the x direction ranges from 20 to 300 nanometer. The amount of thermal transport of this heterobilayer is predicted to be 19.95 W m−1 K−1 over an unlimited length. In this work, the van der Waals thickness is used to predict this thermal transmission. The length of the nanosheet appears to boost the in-plane heat conduction of the germanene/stanene bilayer. For a better understanding of in-plane thermal conduction, the phonon density of states is determined. The characterization of germanene/stanene nanostructure proposed in this study would give a decent knowledge to make it a promising bilayer for the thermoelectric applications owing to its low thermal conductivity.","PeriodicalId":188366,"journal":{"name":"2021 3rd International Conference on Electrical & Electronic Engineering (ICEEE)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121901597","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 : 2021-12-22DOI: 10.1109/ICEEE54059.2021.9718789
S. M. Mahidul Hasan, Md. Rezwanul Ahsan, Md. Dara Abdus Satter
The Internet of things (IoT) is an arising innovation, which changed the industrialization system at a higher level. Staying away from significant catastrophes in the food business or unexpected trivial issues of noticing temperature, humidity, and duct can bring about combined misfortune in food commerce. The main focus of this research has been on how a strategic distance can be maintained from those business misfortunes by incorporating IoT. In food shops, hazardous foods should be kept at a certain level of temperature, humidity, and satisfactory dust level to avert poisoning bacteria. To accomplish the task, IoT-based sensors are used within this research to collect variations of temperature, humidity, and dust level of any hazardous food’s climate and provide required activities with a precise choice. The proposed temperature, humidity, and dust monitoring system has been tested at AJWAH Bake and Pastry shop. The onsite experimental data shows that the system prototype is very effective in observing the food environment and can be utilized at food shops.
{"title":"IoT-Cloud-Based Low-Cost Temperature, Humidity, and Dust Monitoring System to Prevent Food Poisoning","authors":"S. M. Mahidul Hasan, Md. Rezwanul Ahsan, Md. Dara Abdus Satter","doi":"10.1109/ICEEE54059.2021.9718789","DOIUrl":"https://doi.org/10.1109/ICEEE54059.2021.9718789","url":null,"abstract":"The Internet of things (IoT) is an arising innovation, which changed the industrialization system at a higher level. Staying away from significant catastrophes in the food business or unexpected trivial issues of noticing temperature, humidity, and duct can bring about combined misfortune in food commerce. The main focus of this research has been on how a strategic distance can be maintained from those business misfortunes by incorporating IoT. In food shops, hazardous foods should be kept at a certain level of temperature, humidity, and satisfactory dust level to avert poisoning bacteria. To accomplish the task, IoT-based sensors are used within this research to collect variations of temperature, humidity, and dust level of any hazardous food’s climate and provide required activities with a precise choice. The proposed temperature, humidity, and dust monitoring system has been tested at AJWAH Bake and Pastry shop. The onsite experimental data shows that the system prototype is very effective in observing the food environment and can be utilized at food shops.","PeriodicalId":188366,"journal":{"name":"2021 3rd International Conference on Electrical & Electronic Engineering (ICEEE)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124415206","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 : 2021-12-22DOI: 10.1109/ICEEE54059.2021.9718776
Md. Farukuzzaman Faruk
Coronavirus illness, commonly abbreviated as COVID-19, has been designated a global pandemic. To prevent the spread of this deadly virus, those who are infected must be quarantined or evacuated. In this situation, a quick and systematic testing toolkit is required. Recent research has discovered that radiography chest CT has significant patterns and attributes that may be utilized to precisely identify COVID-19. A deep learning-based network called ResidualCovid-Net was suggested in this study to identify COVID-19 infestations using CT scans. The proposed ResidualCovid-Net is inspired by the original Resnet architecture. Another barrier in this aspect is clinically distinguishing among COVID-19, pneumonia and normal instances. ResidualCovid-Net was designed to identify anomalies in CT scans that may successfully delineate COVID-19, common pneumonia and normal cases. Gradients weighted class activation maps showed how well the network located anomalies in CT images and demonstrated the network’s generalization ability.
{"title":"ResidualCovid-Net: An Interpretable Deep Network to Screen COVID-19 Utilizing Chest CT Images","authors":"Md. Farukuzzaman Faruk","doi":"10.1109/ICEEE54059.2021.9718776","DOIUrl":"https://doi.org/10.1109/ICEEE54059.2021.9718776","url":null,"abstract":"Coronavirus illness, commonly abbreviated as COVID-19, has been designated a global pandemic. To prevent the spread of this deadly virus, those who are infected must be quarantined or evacuated. In this situation, a quick and systematic testing toolkit is required. Recent research has discovered that radiography chest CT has significant patterns and attributes that may be utilized to precisely identify COVID-19. A deep learning-based network called ResidualCovid-Net was suggested in this study to identify COVID-19 infestations using CT scans. The proposed ResidualCovid-Net is inspired by the original Resnet architecture. Another barrier in this aspect is clinically distinguishing among COVID-19, pneumonia and normal instances. ResidualCovid-Net was designed to identify anomalies in CT scans that may successfully delineate COVID-19, common pneumonia and normal cases. Gradients weighted class activation maps showed how well the network located anomalies in CT images and demonstrated the network’s generalization ability.","PeriodicalId":188366,"journal":{"name":"2021 3rd International Conference on Electrical & Electronic Engineering (ICEEE)","volume":"132 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131481226","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 : 2021-12-22DOI: 10.1109/ICEEE54059.2021.9719000
Md Abdur Raiyan, S. C. Mohonta
In Brain Computer Interface (BCI), for precise prediction of brain activity, it is important to know which part of the brain is responsible for which activity. Electroencephalography (EEG) signal which conveys the information of such brain activity is recorded using a number of electrodes from all over the skull. In this study, a comparison from a machine learning perspective has been made to investigate which sets of electrodes that mean which part of the brain shows more neural activity during execution or imagination of fist movement. Here, all the preprocessing steps have been done using EEGLAB on MATLAB, and the normalized band powers of five brain rhythms such as alpha, beta, gamma, delta and theta have been used as features. Finally, a supervised machine learning technique – Support Vector Machine (SVM) has been implemented which took those features as input for classification. This study shows that the channel set with more electrodes can distinguish between executed and imaginary fist movement more accurately. Therefore, these findings can be used to understand brain functionality more distinctly and be applied to predict motor movement more precisely in future BCI research.
{"title":"Comparative Study on EEG Based Motor Movement Classification Using Different Sets of Electrode Channels","authors":"Md Abdur Raiyan, S. C. Mohonta","doi":"10.1109/ICEEE54059.2021.9719000","DOIUrl":"https://doi.org/10.1109/ICEEE54059.2021.9719000","url":null,"abstract":"In Brain Computer Interface (BCI), for precise prediction of brain activity, it is important to know which part of the brain is responsible for which activity. Electroencephalography (EEG) signal which conveys the information of such brain activity is recorded using a number of electrodes from all over the skull. In this study, a comparison from a machine learning perspective has been made to investigate which sets of electrodes that mean which part of the brain shows more neural activity during execution or imagination of fist movement. Here, all the preprocessing steps have been done using EEGLAB on MATLAB, and the normalized band powers of five brain rhythms such as alpha, beta, gamma, delta and theta have been used as features. Finally, a supervised machine learning technique – Support Vector Machine (SVM) has been implemented which took those features as input for classification. This study shows that the channel set with more electrodes can distinguish between executed and imaginary fist movement more accurately. Therefore, these findings can be used to understand brain functionality more distinctly and be applied to predict motor movement more precisely in future BCI research.","PeriodicalId":188366,"journal":{"name":"2021 3rd International Conference on Electrical & Electronic Engineering (ICEEE)","volume":"106 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134343065","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 : 2021-12-22DOI: 10.1109/ICEEE54059.2021.9718777
Abir Ebna Harun, Mohammad Ashfak Habib
Urinalysis is a common medical test that can be costly and inconvenient in medical facilities. The use of point-of-care(POC) test devices, smartphones, manifolds, and other additional tools can make urinalysis easier in a home-based environment. In this paper, we are proposing a new system that can be used to performing a laboratory-free urinalysis with the help of a urine test strip and a smartphone device. Our system contains several image pre-processing steps and an artificial neural network mapping model to analyze the color pixels of the urine test strip. By following our proposed solution, the user can acquire an accurate computer vision integrated urinalysis result.
{"title":"An Alternate Solution for Smartphone-Based Urinalysis","authors":"Abir Ebna Harun, Mohammad Ashfak Habib","doi":"10.1109/ICEEE54059.2021.9718777","DOIUrl":"https://doi.org/10.1109/ICEEE54059.2021.9718777","url":null,"abstract":"Urinalysis is a common medical test that can be costly and inconvenient in medical facilities. The use of point-of-care(POC) test devices, smartphones, manifolds, and other additional tools can make urinalysis easier in a home-based environment. In this paper, we are proposing a new system that can be used to performing a laboratory-free urinalysis with the help of a urine test strip and a smartphone device. Our system contains several image pre-processing steps and an artificial neural network mapping model to analyze the color pixels of the urine test strip. By following our proposed solution, the user can acquire an accurate computer vision integrated urinalysis result.","PeriodicalId":188366,"journal":{"name":"2021 3rd International Conference on Electrical & Electronic Engineering (ICEEE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130336778","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 : 2021-12-22DOI: 10.1109/ICEEE54059.2021.9718800
Mahib Tanvir, M. Alam, Dipanwita Saha, Shahid A. Hasib, S. Islam
Sign Language is the elementary communication media for Deaf & Mute (D&M) people. On the other hand, it seems too tenacious for the general people to understand this language. In order to tear out this communication barrier, a real-time automated translator is essential. Through this research, a computer vision-based approach has been developed for the recognition of Bangla Sign Language (BdSL) characters. In this work, a deep learning-based recognition model has been developed. Adaptive thresholding has been integrated with 2D Convolutional Neural Network (CNN) to construct this model. Proposed model has been trained to build this real-time automated translator through our own created dataset (dataset containing 3600 different images for 36 distinct characters). The proposed model has been trained and tested with 2880 (80%) training images and 720 (20%) testing images respectively. Thirty-six unique characters of Bangla Sign Language can be recognized through this model with significant accuracy. The model delivers validation accuracy of 99.72% and validation loss of 0.73%. A significant result has been achieved for the recognition and translation of Bangla Sign Language characters with this dataset over other existing Bangla Sign Language Recognition model.
{"title":"Real-Time Recognition of Bangla Sign Language Characters: A Computer Vision Based Approach Using Convolutional Neural Network","authors":"Mahib Tanvir, M. Alam, Dipanwita Saha, Shahid A. Hasib, S. Islam","doi":"10.1109/ICEEE54059.2021.9718800","DOIUrl":"https://doi.org/10.1109/ICEEE54059.2021.9718800","url":null,"abstract":"Sign Language is the elementary communication media for Deaf & Mute (D&M) people. On the other hand, it seems too tenacious for the general people to understand this language. In order to tear out this communication barrier, a real-time automated translator is essential. Through this research, a computer vision-based approach has been developed for the recognition of Bangla Sign Language (BdSL) characters. In this work, a deep learning-based recognition model has been developed. Adaptive thresholding has been integrated with 2D Convolutional Neural Network (CNN) to construct this model. Proposed model has been trained to build this real-time automated translator through our own created dataset (dataset containing 3600 different images for 36 distinct characters). The proposed model has been trained and tested with 2880 (80%) training images and 720 (20%) testing images respectively. Thirty-six unique characters of Bangla Sign Language can be recognized through this model with significant accuracy. The model delivers validation accuracy of 99.72% and validation loss of 0.73%. A significant result has been achieved for the recognition and translation of Bangla Sign Language characters with this dataset over other existing Bangla Sign Language Recognition model.","PeriodicalId":188366,"journal":{"name":"2021 3rd International Conference on Electrical & Electronic Engineering (ICEEE)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130396357","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 : 2021-12-22DOI: 10.1109/ICEEE54059.2021.9718773
M. A. Motin, Partha Pratim Das, C. Karmakar, M. Palaniswami
Regular monitoring of vital signs is an effective way to prevent life threatening health hazards. In this regard, wearable healthcare devices can play a significant role in the pervasive monitoring of vital signs. In this context, we propose a compact fingertip reflective type pulse oximeter prototype using low cost integrated circuit as a wearable healthcare device for monitoring vital signs. In addition, this device is not only easy to monitor but also suitable for long time monitoring. In this prototype, blood oxygen saturation (SpO2) and heart rate (HR) are measured for home and mobile applications. A complete system design and its evaluation with detailed real time processing scenario are demonstrated in this paper. The test platform is completed by PC based online and offline data analysis. Android interface is also presented for smartphone based online healthcare monitoring.
{"title":"Compact Pulse Oximeter Designed for Blood Oxygen Saturation and Heart Rate Monitoring","authors":"M. A. Motin, Partha Pratim Das, C. Karmakar, M. Palaniswami","doi":"10.1109/ICEEE54059.2021.9718773","DOIUrl":"https://doi.org/10.1109/ICEEE54059.2021.9718773","url":null,"abstract":"Regular monitoring of vital signs is an effective way to prevent life threatening health hazards. In this regard, wearable healthcare devices can play a significant role in the pervasive monitoring of vital signs. In this context, we propose a compact fingertip reflective type pulse oximeter prototype using low cost integrated circuit as a wearable healthcare device for monitoring vital signs. In addition, this device is not only easy to monitor but also suitable for long time monitoring. In this prototype, blood oxygen saturation (SpO2) and heart rate (HR) are measured for home and mobile applications. A complete system design and its evaluation with detailed real time processing scenario are demonstrated in this paper. The test platform is completed by PC based online and offline data analysis. Android interface is also presented for smartphone based online healthcare monitoring.","PeriodicalId":188366,"journal":{"name":"2021 3rd International Conference on Electrical & Electronic Engineering (ICEEE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130493315","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：