In this work, we consider a performance comparison of deep learning-based approaches to the problem of spectrum sensing (SS) in cognitive radios. Towards this end, we use signal processing (SP) features such as energy, differential entropy, geometric power and p-norm. For the classification problem of SS, we employ deep neural network (NN) architectures such as multi-layer perceptron (MLP), convolutional NN, fully convolutional network, residual NN (ResNet), long short-term memory and temporal convolutional network. Through extensive experiments based on real-world captured datasets and Monte Carlo simulations, we show that MLP and ResNet architectures offer the best performance in terms of probability of detection, for a given predefined level of probability of false-alarm. Further, we show that NN architectures trained with a combined set of the SP features yield the best performance.
{"title":"Deep Neural Network Architectures for Spectrum Sensing Using Signal Processing Features","authors":"Shreeram Suresh Chandra, Akshay Upadhye, Purushothaman Saravanan, Sanjeev Gurugopinath, R. Muralishankar","doi":"10.1109/DISCOVER52564.2021.9663583","DOIUrl":"https://doi.org/10.1109/DISCOVER52564.2021.9663583","url":null,"abstract":"In this work, we consider a performance comparison of deep learning-based approaches to the problem of spectrum sensing (SS) in cognitive radios. Towards this end, we use signal processing (SP) features such as energy, differential entropy, geometric power and p-norm. For the classification problem of SS, we employ deep neural network (NN) architectures such as multi-layer perceptron (MLP), convolutional NN, fully convolutional network, residual NN (ResNet), long short-term memory and temporal convolutional network. Through extensive experiments based on real-world captured datasets and Monte Carlo simulations, we show that MLP and ResNet architectures offer the best performance in terms of probability of detection, for a given predefined level of probability of false-alarm. Further, we show that NN architectures trained with a combined set of the SP features yield the best performance.","PeriodicalId":413789,"journal":{"name":"2021 IEEE International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics (DISCOVER)","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125914110","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 : 2021-11-19DOI: 10.1109/DISCOVER52564.2021.9663697
Prashanth Barla, Vinod Kumar Joshi, S. Bhat
Among all spintronics devices, spin transfer torque (STT) magnetic tunnel junction (MTJ) is the most promising candidate for logic-in-memory (LIM) architecture. It alleviates the performance degradation observed in the present CMOS circuits which are built using standard von-Neumann architecture. However STT-MTJ suffers the issues such as switching delay due to stochasticity as well as wastage of write power. Hence, in this work continuous monitoring and self-write-termination (SWT) process is adopted for STT-MTJs and studied the performance of all the logic gates; AND/NAND, OR/NOR and XOR/XNOR developed using LIM architecture. Investigation of the read/write power, read/write delay, read/write power delay product and transistor count of SWT-STT-MTJ/CMOS logic gates are performed and compared them with its conventional counterparts. Further, Monte-Carlo simulations are also conducted to study the behavior of hybrid logic gates for variations that could occur during fabrication. The simulation results reveal that SWT-STT-MTJ/CMOS logic gates dissipates lower power, PDP and produce quicker output response.
{"title":"Design and Analysis of Self-write-terminated Hybrid STT-MTJ/CMOS Logic Gates using LIM Architecture","authors":"Prashanth Barla, Vinod Kumar Joshi, S. Bhat","doi":"10.1109/DISCOVER52564.2021.9663697","DOIUrl":"https://doi.org/10.1109/DISCOVER52564.2021.9663697","url":null,"abstract":"Among all spintronics devices, spin transfer torque (STT) magnetic tunnel junction (MTJ) is the most promising candidate for logic-in-memory (LIM) architecture. It alleviates the performance degradation observed in the present CMOS circuits which are built using standard von-Neumann architecture. However STT-MTJ suffers the issues such as switching delay due to stochasticity as well as wastage of write power. Hence, in this work continuous monitoring and self-write-termination (SWT) process is adopted for STT-MTJs and studied the performance of all the logic gates; AND/NAND, OR/NOR and XOR/XNOR developed using LIM architecture. Investigation of the read/write power, read/write delay, read/write power delay product and transistor count of SWT-STT-MTJ/CMOS logic gates are performed and compared them with its conventional counterparts. Further, Monte-Carlo simulations are also conducted to study the behavior of hybrid logic gates for variations that could occur during fabrication. The simulation results reveal that SWT-STT-MTJ/CMOS logic gates dissipates lower power, PDP and produce quicker output response.","PeriodicalId":413789,"journal":{"name":"2021 IEEE International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics (DISCOVER)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115049574","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 : 2021-11-19DOI: 10.1109/DISCOVER52564.2021.9663579
Nune Divya, B. H. Chandana, D. Harika
In this paper, the implementation of digital receiver using Intel IPP (Integrated Performance Primitives) library functions with digital filtering approach via fast convolution method is discussed using Xeon-processor, it can simulate more number of generated samples at the receiver using performance primitives by intel. The proposed work examines various benchmark functions in Intel IPP library to compute matched filter responses using time and frequency domain responses. Intel IPP executes multiple programs using SIMD (Single-Instruction Multiple Data) stream along with reducing computational time, cost and prolonged processor life time.
{"title":"Implementation of Radar Digital Receiver based on Xeon-Processor using Intel IPP","authors":"Nune Divya, B. H. Chandana, D. Harika","doi":"10.1109/DISCOVER52564.2021.9663579","DOIUrl":"https://doi.org/10.1109/DISCOVER52564.2021.9663579","url":null,"abstract":"In this paper, the implementation of digital receiver using Intel IPP (Integrated Performance Primitives) library functions with digital filtering approach via fast convolution method is discussed using Xeon-processor, it can simulate more number of generated samples at the receiver using performance primitives by intel. The proposed work examines various benchmark functions in Intel IPP library to compute matched filter responses using time and frequency domain responses. Intel IPP executes multiple programs using SIMD (Single-Instruction Multiple Data) stream along with reducing computational time, cost and prolonged processor life time.","PeriodicalId":413789,"journal":{"name":"2021 IEEE International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics (DISCOVER)","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124027056","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 : 2021-11-19DOI: 10.1109/DISCOVER52564.2021.9663705
Archana Yashodhar, Shashidhar Kini
One of the neurogenerative disorders affected by many adults is Alzheimer’s Disease (AD). Disease prediction is also a difficult task. Except in the healthcare domain, the principles of artificial learning and data processing are commonly used. In this research paper, using the Machine Learning method, we applied various classification algorithms on the data sets and concluded which algorithms provide the best results. Also, we have used different possible methods to evaluate the model.
{"title":"Performance Measurements of different Classification techniques for the Alzheimer’s Disease Neuroimaging Initiative","authors":"Archana Yashodhar, Shashidhar Kini","doi":"10.1109/DISCOVER52564.2021.9663705","DOIUrl":"https://doi.org/10.1109/DISCOVER52564.2021.9663705","url":null,"abstract":"One of the neurogenerative disorders affected by many adults is Alzheimer’s Disease (AD). Disease prediction is also a difficult task. Except in the healthcare domain, the principles of artificial learning and data processing are commonly used. In this research paper, using the Machine Learning method, we applied various classification algorithms on the data sets and concluded which algorithms provide the best results. Also, we have used different possible methods to evaluate the model.","PeriodicalId":413789,"journal":{"name":"2021 IEEE International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics (DISCOVER)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131540300","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 : 2021-11-19DOI: 10.1109/DISCOVER52564.2021.9663631
Annapurna Anant Shanbhag, Chinmai Shetty, A. Ananth, Anjali Shridhar Shetty, K. Kavanashree Nayak, B. R. Rakshitha
Heart Attack is one of the most common diseases observed in people of middle age as well as old age in the present day scenario. This may be due to unhealthy food habits and negligence of health in most people. Detecting the risk of heart attack and taking timely medication, can prevent serious illness. In this paper we explain about the different machine learning approaches and techniques used for predicting the probability of heart-attack risk. Different models are applied for heart-attack risk prediction. The probability of heart attack risk is displayed through a website. If a person is found having risk, suitable precautions are displayed under the guidance of the cardiologist. The proposed work analyses whether the person has a normal range of values for some highly contributing attributes which lead to heart attack like Cholesterol, Blood pressure, Blood sugar. The proposed work has better results compared to the previous work in terms of accuracy of prediction with highest value of accuracy as 85.7% for SVM model.
{"title":"Heart Attack Probability Analysis Using Machine Learning","authors":"Annapurna Anant Shanbhag, Chinmai Shetty, A. Ananth, Anjali Shridhar Shetty, K. Kavanashree Nayak, B. R. Rakshitha","doi":"10.1109/DISCOVER52564.2021.9663631","DOIUrl":"https://doi.org/10.1109/DISCOVER52564.2021.9663631","url":null,"abstract":"Heart Attack is one of the most common diseases observed in people of middle age as well as old age in the present day scenario. This may be due to unhealthy food habits and negligence of health in most people. Detecting the risk of heart attack and taking timely medication, can prevent serious illness. In this paper we explain about the different machine learning approaches and techniques used for predicting the probability of heart-attack risk. Different models are applied for heart-attack risk prediction. The probability of heart attack risk is displayed through a website. If a person is found having risk, suitable precautions are displayed under the guidance of the cardiologist. The proposed work analyses whether the person has a normal range of values for some highly contributing attributes which lead to heart attack like Cholesterol, Blood pressure, Blood sugar. The proposed work has better results compared to the previous work in terms of accuracy of prediction with highest value of accuracy as 85.7% for SVM model.","PeriodicalId":413789,"journal":{"name":"2021 IEEE International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics (DISCOVER)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131622971","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 : 2021-11-19DOI: 10.1109/DISCOVER52564.2021.9663722
N. Sreenivasa, B. A. Mohan, E. G. Satish, Roshan Fernandes, P. Ramesh Naidu, T. Vinay
Automation plays key role in our day today lives making our life easier, simpler and easy living. In this research paper a prototype for smart office automation is designed and implemented. This project includes subsystems like managing energy saving, security and alarming systems. The sensors and actuators are used to mine the real time data from the indoor office environment. All the sensors and actuators are connected to the embedded system link microcontroller unit. It further processes the data, analyze and control the subsystems in office environment. The electrical/electronic devices such as bulbs, fans, buzzer, sensors like temperature, humidity and motion sensors are connected to the microcontroller, which will generate values/data when they cross the certain threshold values. This office system facilitate control of electrical/electronic devices like door-access, illuminating, lighting and ventilating system in order to save energy and uplifts employees’ and customers satisfactions. This work promotes indoor office automation along with saving the cost for the employer and providing comforts to the employees.
{"title":"Prototyping of Intelligent Office Monitoring and Control System Using IoT","authors":"N. Sreenivasa, B. A. Mohan, E. G. Satish, Roshan Fernandes, P. Ramesh Naidu, T. Vinay","doi":"10.1109/DISCOVER52564.2021.9663722","DOIUrl":"https://doi.org/10.1109/DISCOVER52564.2021.9663722","url":null,"abstract":"Automation plays key role in our day today lives making our life easier, simpler and easy living. In this research paper a prototype for smart office automation is designed and implemented. This project includes subsystems like managing energy saving, security and alarming systems. The sensors and actuators are used to mine the real time data from the indoor office environment. All the sensors and actuators are connected to the embedded system link microcontroller unit. It further processes the data, analyze and control the subsystems in office environment. The electrical/electronic devices such as bulbs, fans, buzzer, sensors like temperature, humidity and motion sensors are connected to the microcontroller, which will generate values/data when they cross the certain threshold values. This office system facilitate control of electrical/electronic devices like door-access, illuminating, lighting and ventilating system in order to save energy and uplifts employees’ and customers satisfactions. This work promotes indoor office automation along with saving the cost for the employer and providing comforts to the employees.","PeriodicalId":413789,"journal":{"name":"2021 IEEE International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics (DISCOVER)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133285633","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 : 2021-11-19DOI: 10.1109/DISCOVER52564.2021.9663715
Vamsi Vijay Mohan Dattada, Sreedevi Sasidharan, A. Højlund, K. S. Sridharan
Deep Brain Stimulation (DBS) is an established and effective neuromodulation technique preferred in treating several neurological and neuropsychiatric disorders such as Parkinson’s Disease(PD), epilepsy, obsessive compulsive disorder, depression and several such disorders. Magnetoencephalography (MEG) is a widely used neuroimaging strategy to understand the pathology and the therapeutic effects of DBS in clinical cohorts. One of the significant limitations is the inability to differentiate the DBS stimulation artefact from actual neuronal excitations, especially in lower frequency bands of interest where sub-harmonics of DBS artefacts may obscure the biological response and is a confounder. The primary objective of this study is to understand how DBS stimulation artefacts affect MEG signals and to this end, we employ a phantom based on a water melon. Using this phantom, we record the spectral signature of the DBS stimulation artefact at various DBS frequencies and stimulation voltages, the effect of standard artefact rejection approaches like spatiotemporal signal space separation (tSSS). We present in this paper the results of the initial analysis.
{"title":"How Does Deep Brain Stimulation Affect Magnetoencephalography Data?","authors":"Vamsi Vijay Mohan Dattada, Sreedevi Sasidharan, A. Højlund, K. S. Sridharan","doi":"10.1109/DISCOVER52564.2021.9663715","DOIUrl":"https://doi.org/10.1109/DISCOVER52564.2021.9663715","url":null,"abstract":"Deep Brain Stimulation (DBS) is an established and effective neuromodulation technique preferred in treating several neurological and neuropsychiatric disorders such as Parkinson’s Disease(PD), epilepsy, obsessive compulsive disorder, depression and several such disorders. Magnetoencephalography (MEG) is a widely used neuroimaging strategy to understand the pathology and the therapeutic effects of DBS in clinical cohorts. One of the significant limitations is the inability to differentiate the DBS stimulation artefact from actual neuronal excitations, especially in lower frequency bands of interest where sub-harmonics of DBS artefacts may obscure the biological response and is a confounder. The primary objective of this study is to understand how DBS stimulation artefacts affect MEG signals and to this end, we employ a phantom based on a water melon. Using this phantom, we record the spectral signature of the DBS stimulation artefact at various DBS frequencies and stimulation voltages, the effect of standard artefact rejection approaches like spatiotemporal signal space separation (tSSS). We present in this paper the results of the initial analysis.","PeriodicalId":413789,"journal":{"name":"2021 IEEE International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics (DISCOVER)","volume":"117 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129589399","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 : 2021-11-19DOI: 10.1109/DISCOVER52564.2021.9663704
Rumana Ali, Vinayambika S. Bhat, C. Abhishek, Akash Joseph, Mohammed Arfadh, Akarsh Manoj
The paper focuses on harvesting energy from an energy source consisting of electrolytic solution and electrodes. The design focuses on using different ionized chemicals as electrolytes, which includes NaCl in Faucet water, Vinegar and Electrodes including Zinc, Copper, Magnesium, Graphite, Aluminium. The cells are arranged in cascade and parallel setup, and the voltage, current values are noted and studied using a Multimeter. Results depict that the energy produced from different electrolytes and electrode combinations vary based on the reactivity of the element, which can be referred from the electrochemical series. Alongside the thought of the device operation, aspect, overall expenditure a six-celled Graphite Magnesium Electrolytic-cell battery in a cascade configuration, using NaCl-water-electrolyte produced a voltage of 9.0 volts for seventeen hours further can be improvised by changing electrolyte. The device designed can be used to activate an LED lamp.
{"title":"Novel Approach to Harvest Energy from Salinity Gradient","authors":"Rumana Ali, Vinayambika S. Bhat, C. Abhishek, Akash Joseph, Mohammed Arfadh, Akarsh Manoj","doi":"10.1109/DISCOVER52564.2021.9663704","DOIUrl":"https://doi.org/10.1109/DISCOVER52564.2021.9663704","url":null,"abstract":"The paper focuses on harvesting energy from an energy source consisting of electrolytic solution and electrodes. The design focuses on using different ionized chemicals as electrolytes, which includes NaCl in Faucet water, Vinegar and Electrodes including Zinc, Copper, Magnesium, Graphite, Aluminium. The cells are arranged in cascade and parallel setup, and the voltage, current values are noted and studied using a Multimeter. Results depict that the energy produced from different electrolytes and electrode combinations vary based on the reactivity of the element, which can be referred from the electrochemical series. Alongside the thought of the device operation, aspect, overall expenditure a six-celled Graphite Magnesium Electrolytic-cell battery in a cascade configuration, using NaCl-water-electrolyte produced a voltage of 9.0 volts for seventeen hours further can be improvised by changing electrolyte. The device designed can be used to activate an LED lamp.","PeriodicalId":413789,"journal":{"name":"2021 IEEE International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics (DISCOVER)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129951636","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 : 2021-11-19DOI: 10.1109/DISCOVER52564.2021.9663596
Praahas Amin, A. Khan
Myoelectric control systems are gaining popularity with the availability of commercial, low-cost, surface electromyography sensors. These systems can be used for gesture recognition which finds application in human-machine interfaces. The gestures are recognized using pattern recognition algorithms. Machine learning or deep learning techniques can be applied for the classification of gestures. In this paper, a user-specific 1-Dimensional Convolution Neural Network is proposed for the classification of Surface Electromyography data recorded using a commercially available surface electromyography recording device to perform offline classification of 5 hand gestures using limited data of less than 400 samples. An average accuracy of 82%±3% was achieved during the study after cross-validation of the data using 5-fold stratified cross-validation.
{"title":"A Study on the Application of One Dimension Convolutional Neural Network for Classification of Gestures from Surface Electromyography Data","authors":"Praahas Amin, A. Khan","doi":"10.1109/DISCOVER52564.2021.9663596","DOIUrl":"https://doi.org/10.1109/DISCOVER52564.2021.9663596","url":null,"abstract":"Myoelectric control systems are gaining popularity with the availability of commercial, low-cost, surface electromyography sensors. These systems can be used for gesture recognition which finds application in human-machine interfaces. The gestures are recognized using pattern recognition algorithms. Machine learning or deep learning techniques can be applied for the classification of gestures. In this paper, a user-specific 1-Dimensional Convolution Neural Network is proposed for the classification of Surface Electromyography data recorded using a commercially available surface electromyography recording device to perform offline classification of 5 hand gestures using limited data of less than 400 samples. An average accuracy of 82%±3% was achieved during the study after cross-validation of the data using 5-fold stratified cross-validation.","PeriodicalId":413789,"journal":{"name":"2021 IEEE International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics (DISCOVER)","volume":"2014 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132263792","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 : 2021-11-19DOI: 10.1109/DISCOVER52564.2021.9663661
R. Vandana, P. S. Venugopala, B. Ashwini
In the modern world, education system has reached a new destination due to the introduction of concept called “smart classroom”. However, when we are speaking about any classroom the attendance system still remains primitive. The traditional attendance system where the teacher/lecturer calls the name of students to mark their attendance in an attendance register is a manual method which is found to be not suitable for a smart class due to a list of disadvantages. The automatic attendance management will replace the manual method. This dynamic attendance management system will consider the physiological features of the human beings for uniquely identifying them. Hence we are using a biometric based attendance system. There are many biometric processes, among which face recognition is the best method. In the proposed project, we are going to describe the attendance without human interference. In this method a camera, fixed within the classroom will capture the image, the faces are detected and then they are compared with the faces in the database and finally the attendance is marked. It also proposes a single image-based face liveness detection method for discriminating 2-D paper masks from the live faces. Freely available machine learning and deep learning tools like dlib, Keras are used for making the face recognition faster and accurate one. This makes the system suitable in a real life scenario.
{"title":"Neural Network based Biometric Attendance System","authors":"R. Vandana, P. S. Venugopala, B. Ashwini","doi":"10.1109/DISCOVER52564.2021.9663661","DOIUrl":"https://doi.org/10.1109/DISCOVER52564.2021.9663661","url":null,"abstract":"In the modern world, education system has reached a new destination due to the introduction of concept called “smart classroom”. However, when we are speaking about any classroom the attendance system still remains primitive. The traditional attendance system where the teacher/lecturer calls the name of students to mark their attendance in an attendance register is a manual method which is found to be not suitable for a smart class due to a list of disadvantages. The automatic attendance management will replace the manual method. This dynamic attendance management system will consider the physiological features of the human beings for uniquely identifying them. Hence we are using a biometric based attendance system. There are many biometric processes, among which face recognition is the best method. In the proposed project, we are going to describe the attendance without human interference. In this method a camera, fixed within the classroom will capture the image, the faces are detected and then they are compared with the faces in the database and finally the attendance is marked. It also proposes a single image-based face liveness detection method for discriminating 2-D paper masks from the live faces. Freely available machine learning and deep learning tools like dlib, Keras are used for making the face recognition faster and accurate one. This makes the system suitable in a real life scenario.","PeriodicalId":413789,"journal":{"name":"2021 IEEE International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics (DISCOVER)","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126693648","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：