Pub Date : 2022-12-07DOI: 10.1109/ICETECC56662.2022.10069504
Sanaullah, Rumina Nawab Ali, Muhammad Farrukh Shahid
A person’s present state of mind is determined by a complex collection of brain activities that make up their mental state. It is influenced by several internal and external aspects of the brain. By examining an individual’s EEG patterns, one can ascertain their mental state. In order to recognise and alter harmful or troubling thinking patterns that have a detrimental impact on behaviour and emotions, we classified three different states as: relaxed, neutral, and focused. To classify and predict the behaviour of a person based on certain mental states, we deployed popular machine learning models like k-NN, RF, XGBOOST, and EL to classify different mental states. Moreover, to predict the mental states, we implemented deep learning models like CNN, RNN, and LSTM. XGBoost achieves the highest classification accuracy (97.29%) with 5-fold cross validation. For the prediction, RNN achieved the highest prediction accuracy of 97.84%.
{"title":"An Novel Approach to Predict and Classify the Mental State of Person using EEG-based Brain-Computer Interface","authors":"Sanaullah, Rumina Nawab Ali, Muhammad Farrukh Shahid","doi":"10.1109/ICETECC56662.2022.10069504","DOIUrl":"https://doi.org/10.1109/ICETECC56662.2022.10069504","url":null,"abstract":"A person’s present state of mind is determined by a complex collection of brain activities that make up their mental state. It is influenced by several internal and external aspects of the brain. By examining an individual’s EEG patterns, one can ascertain their mental state. In order to recognise and alter harmful or troubling thinking patterns that have a detrimental impact on behaviour and emotions, we classified three different states as: relaxed, neutral, and focused. To classify and predict the behaviour of a person based on certain mental states, we deployed popular machine learning models like k-NN, RF, XGBOOST, and EL to classify different mental states. Moreover, to predict the mental states, we implemented deep learning models like CNN, RNN, and LSTM. XGBoost achieves the highest classification accuracy (97.29%) with 5-fold cross validation. For the prediction, RNN achieved the highest prediction accuracy of 97.84%.","PeriodicalId":364463,"journal":{"name":"2022 International Conference on Emerging Technologies in Electronics, Computing and Communication (ICETECC)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126387536","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}
This paper proposes a cost-effective way for the object detection and classification of objects modeled as 3D renders, via Deep Learning. 3D modeling is the process of manipulating edges, vertices, and polygons in artificial 3D space that creates mathematical coordinated representations of the surface. In this research, we propose to use a stereo camera and a 2D laser scanner (LiDAR) for the construction of 3D object models. We created a 3D model of an object using a stereo camera. Video of objects was captured maintaining the right angles all the time. Then with the help of Intel Real Sense Viewer, a 3D polygon mesh was created, which was converted to a point cloud. A two-dimensional (2D) laser scanner was used to make several chunks of 2D scans from various sides of the object. We then fused the point cloud of the obtained chunks to build a 3D model. We then combined the point clouds obtained from both sources using the Iterative Closest Point (ICP) algorithm. The fused point cloud resulted in the formation of a denser and crispier dataset to be used for Deep Learning. The aforementioned deep learning algorithm, Point Net, encodes sparse point cloud data efficiently and shows very strong performance on par with the state of the art. We have formed a dataset using stereo camera, LIDAR and ICP among which we have obtained the highest accuracy results from ICP algorithm dataset.
本文提出了一种经济有效的方法,通过深度学习对3D渲染建模的对象进行检测和分类。3D建模是在人工3D空间中操作边缘、顶点和多边形的过程,从而创建表面的数学协调表示。在本研究中,我们建议使用立体相机和二维激光扫描仪(LiDAR)来构建三维物体模型。我们用立体摄像机创建了一个物体的3D模型。拍摄到的物体的视频一直保持着正确的角度。然后在Intel Real Sense Viewer的帮助下,创建了一个3D多边形网格,并将其转换为点云。一个二维激光扫描仪被用来从物体的不同侧面进行几块二维扫描。然后,我们将得到的块的点云融合,以建立一个三维模型。然后,我们使用迭代最近点(ICP)算法将从两个来源获得的点云结合起来。融合的点云形成了一个更密集、更清晰的数据集,用于深度学习。前面提到的深度学习算法Point Net有效地编码了稀疏的点云数据,并显示出与当前技术水平相当的强大性能。我们使用立体相机、激光雷达和ICP组成了一个数据集,其中ICP算法数据集获得了精度最高的结果。
{"title":"Object Detection from 3D Point Cloud Using Deep Learning (SFM)","authors":"Hareem Rizvi, Nimra Zahoor Qazi, Talha Shakil, Asad-ur-Rehman, Yawar Rehman","doi":"10.1109/ICETECC56662.2022.10069939","DOIUrl":"https://doi.org/10.1109/ICETECC56662.2022.10069939","url":null,"abstract":"This paper proposes a cost-effective way for the object detection and classification of objects modeled as 3D renders, via Deep Learning. 3D modeling is the process of manipulating edges, vertices, and polygons in artificial 3D space that creates mathematical coordinated representations of the surface. In this research, we propose to use a stereo camera and a 2D laser scanner (LiDAR) for the construction of 3D object models. We created a 3D model of an object using a stereo camera. Video of objects was captured maintaining the right angles all the time. Then with the help of Intel Real Sense Viewer, a 3D polygon mesh was created, which was converted to a point cloud. A two-dimensional (2D) laser scanner was used to make several chunks of 2D scans from various sides of the object. We then fused the point cloud of the obtained chunks to build a 3D model. We then combined the point clouds obtained from both sources using the Iterative Closest Point (ICP) algorithm. The fused point cloud resulted in the formation of a denser and crispier dataset to be used for Deep Learning. The aforementioned deep learning algorithm, Point Net, encodes sparse point cloud data efficiently and shows very strong performance on par with the state of the art. We have formed a dataset using stereo camera, LIDAR and ICP among which we have obtained the highest accuracy results from ICP algorithm dataset.","PeriodicalId":364463,"journal":{"name":"2022 International Conference on Emerging Technologies in Electronics, Computing and Communication (ICETECC)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122797153","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-07DOI: 10.1109/ICETECC56662.2022.10069346
Aaqib Raza, M. H. Baloch, Irfan Ali, Waqas Ali, M. Hassan, Abdul Karim
Electric vehicles (EVs) are increasing day by day across the world. Due to zero CO2 emission and being environmentally friendly, electric vehicles are steadily gaining in popularity. Energy storage and charging systems are one of the main issues that should be removed completely. This paper provides solutions to charging systems with hybrid sources, plug-in hybrid electric vehicles (PHEVs), and all-electric vehicles (EVs). The application of the Internet of things (IoT) and Artificial Intelligence in monitoring the performance of a charging system, and fully autonomous driving electric vehicles by using different sensors connected. A self-charging system can be implemented and the exchange of information between the vehicle and its surroundings. Artificial Intelligence (AI) refers to the human mind that can perform tasks and decision-making like human intelligence through different logic and programs. Artificial Intelligence (AI) accelerates electric vehicles towards automation. In the future, IoT and artificial intelligence-based complete autonomous driving vehicles enable us to reduce battery charging, parking, and traffic issues and change the infrastructure into smart cities.
{"title":"Artificial Intelligence and IoT-Based Autonomous Hybrid Electric Vehicle with Self-Charging Infrastructure","authors":"Aaqib Raza, M. H. Baloch, Irfan Ali, Waqas Ali, M. Hassan, Abdul Karim","doi":"10.1109/ICETECC56662.2022.10069346","DOIUrl":"https://doi.org/10.1109/ICETECC56662.2022.10069346","url":null,"abstract":"Electric vehicles (EVs) are increasing day by day across the world. Due to zero CO2 emission and being environmentally friendly, electric vehicles are steadily gaining in popularity. Energy storage and charging systems are one of the main issues that should be removed completely. This paper provides solutions to charging systems with hybrid sources, plug-in hybrid electric vehicles (PHEVs), and all-electric vehicles (EVs). The application of the Internet of things (IoT) and Artificial Intelligence in monitoring the performance of a charging system, and fully autonomous driving electric vehicles by using different sensors connected. A self-charging system can be implemented and the exchange of information between the vehicle and its surroundings. Artificial Intelligence (AI) refers to the human mind that can perform tasks and decision-making like human intelligence through different logic and programs. Artificial Intelligence (AI) accelerates electric vehicles towards automation. In the future, IoT and artificial intelligence-based complete autonomous driving vehicles enable us to reduce battery charging, parking, and traffic issues and change the infrastructure into smart cities.","PeriodicalId":364463,"journal":{"name":"2022 International Conference on Emerging Technologies in Electronics, Computing and Communication (ICETECC)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123074630","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-07DOI: 10.1109/ICETECC56662.2022.10069860
S. Jamil, Syed Wasi, A. Mahmood, A.R. Rehman
The design, fabrication and performance evaluation of the knee exoskeleton to assist sit-to-stand (STS) motion are presented. The mathematical, CAD and Simscape/MATLAB modeling for the estimation of exoskeleton torque, the factor of safety (FOS) and controller gains are also included. The knee exoskeleton is equipped with two motion sensors. One sensor is meant to detect the intention of motion. This signal is used as a reference trajectory to derive the actuation system. The 2nd sensor is to measure the actual knee joint position which is used as a feedback element to carry out the control mechanism. The exoskeleton is primarily meant for force augmentation by supplementing up to 20% of knee joint torque required, specifically for elderlies and those under rehabilitation. To reduce the total mass to approximately 7 kg, the device frame is made of light-weight aluminum alloy and a worm-gear DC motor is used as the sole actuator. As against commercially available lower limb exoskeletons, our design is simpler, low-cost and easy to use and maintain. It has been tested on able-bodied subjects and has shown the reliability of operation and user comfort. It is expected that its performance for target users, i.e., people with limited sit-to-stand motion capability will produce good results as well. This device can be modified to carry out support for gait and running tasks.
{"title":"Design and Fabrication of Force Augmenting Exoskeleton using Motion Intention Detection","authors":"S. Jamil, Syed Wasi, A. Mahmood, A.R. Rehman","doi":"10.1109/ICETECC56662.2022.10069860","DOIUrl":"https://doi.org/10.1109/ICETECC56662.2022.10069860","url":null,"abstract":"The design, fabrication and performance evaluation of the knee exoskeleton to assist sit-to-stand (STS) motion are presented. The mathematical, CAD and Simscape/MATLAB modeling for the estimation of exoskeleton torque, the factor of safety (FOS) and controller gains are also included. The knee exoskeleton is equipped with two motion sensors. One sensor is meant to detect the intention of motion. This signal is used as a reference trajectory to derive the actuation system. The 2nd sensor is to measure the actual knee joint position which is used as a feedback element to carry out the control mechanism. The exoskeleton is primarily meant for force augmentation by supplementing up to 20% of knee joint torque required, specifically for elderlies and those under rehabilitation. To reduce the total mass to approximately 7 kg, the device frame is made of light-weight aluminum alloy and a worm-gear DC motor is used as the sole actuator. As against commercially available lower limb exoskeletons, our design is simpler, low-cost and easy to use and maintain. It has been tested on able-bodied subjects and has shown the reliability of operation and user comfort. It is expected that its performance for target users, i.e., people with limited sit-to-stand motion capability will produce good results as well. This device can be modified to carry out support for gait and running tasks.","PeriodicalId":364463,"journal":{"name":"2022 International Conference on Emerging Technologies in Electronics, Computing and Communication (ICETECC)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131861971","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-07DOI: 10.1109/ICETECC56662.2022.10069720
Aqib Shafiq, Sheraz Iqbal, Syed Danish Ali, Anis-ur-Rehman, Muhammad Ali, Raja Tahir Iqbal, Mohtasim Usman
For a variety of reasons, electric vehicles (EVs) are becoming more popular. The primary advantage of EVs is reduced pollution from gas emissions. Rising fuel costs and the depletion of fossil fuels are two more issues that need to be addressed. These elements have a greater impact on a Clean Pakistan. EVs are gaining popularity as a means of reducing CO2 emissions from road travel as well as worldwide fossil fuel usage. Electricity needed to charge an electric vehicle’s battery is often obtained from the grid. When EVs are charged by the electrical grid, the system suffers from serious power challenges. To further promote the use of renewable energy resources (RERs) and lower CO2 emissions, certain solar photovoltaic (PV) systems may want to take into account EV charging. This paper will look at a financial and environmental feasibility analysis for the construction of an electric bike charging station powered by solar photovoltaic. Various system design plans are explored in this study, along with their impact on polluted gases emission and economically significant metrics. The proposed approach’s results are compared to those of a charging station that receives a single charge from the grid. Polluted gas emissions, such as CO2, CO, SO2, and NOX, have been greatly reduced when compared to other current approaches. The study should benefit environmentally friendly and commercially profitable renewable energy-based EV charging options.
{"title":"Economic and environmental analysis for different scenarios of grid-connected Solar PV-based EV charging Station facility using Homer Grid","authors":"Aqib Shafiq, Sheraz Iqbal, Syed Danish Ali, Anis-ur-Rehman, Muhammad Ali, Raja Tahir Iqbal, Mohtasim Usman","doi":"10.1109/ICETECC56662.2022.10069720","DOIUrl":"https://doi.org/10.1109/ICETECC56662.2022.10069720","url":null,"abstract":"For a variety of reasons, electric vehicles (EVs) are becoming more popular. The primary advantage of EVs is reduced pollution from gas emissions. Rising fuel costs and the depletion of fossil fuels are two more issues that need to be addressed. These elements have a greater impact on a Clean Pakistan. EVs are gaining popularity as a means of reducing CO2 emissions from road travel as well as worldwide fossil fuel usage. Electricity needed to charge an electric vehicle’s battery is often obtained from the grid. When EVs are charged by the electrical grid, the system suffers from serious power challenges. To further promote the use of renewable energy resources (RERs) and lower CO2 emissions, certain solar photovoltaic (PV) systems may want to take into account EV charging. This paper will look at a financial and environmental feasibility analysis for the construction of an electric bike charging station powered by solar photovoltaic. Various system design plans are explored in this study, along with their impact on polluted gases emission and economically significant metrics. The proposed approach’s results are compared to those of a charging station that receives a single charge from the grid. Polluted gas emissions, such as CO2, CO, SO2, and NOX, have been greatly reduced when compared to other current approaches. The study should benefit environmentally friendly and commercially profitable renewable energy-based EV charging options.","PeriodicalId":364463,"journal":{"name":"2022 International Conference on Emerging Technologies in Electronics, Computing and Communication (ICETECC)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125263067","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-07DOI: 10.1109/ICETECC56662.2022.10069696
Usman Ali, W. T. Toor
This paper presents an efficient regularization scheme for the single image dehazing. The transmission map has been reguarlized to retrieve a dehazed image. Usually, conventional methods try to improve the initial transmission through guided filtering without considering the potential advantage of improving the guidance as well. We have proposed an efficient regularization scheme that jointly optimizes the transmission map and the guidance. Nonconvex energy function is solved by iterative reweighed least squares. As a result, an improved transmission map is obtained that has edges concurrent with the iteratively updated guidance. The regularized transmission map results in better-quality dehazed image which has improved color fidelity and fine details as demonstrated by the experimental results.
{"title":"Mutually Guided Image Dehazing","authors":"Usman Ali, W. T. Toor","doi":"10.1109/ICETECC56662.2022.10069696","DOIUrl":"https://doi.org/10.1109/ICETECC56662.2022.10069696","url":null,"abstract":"This paper presents an efficient regularization scheme for the single image dehazing. The transmission map has been reguarlized to retrieve a dehazed image. Usually, conventional methods try to improve the initial transmission through guided filtering without considering the potential advantage of improving the guidance as well. We have proposed an efficient regularization scheme that jointly optimizes the transmission map and the guidance. Nonconvex energy function is solved by iterative reweighed least squares. As a result, an improved transmission map is obtained that has edges concurrent with the iteratively updated guidance. The regularized transmission map results in better-quality dehazed image which has improved color fidelity and fine details as demonstrated by the experimental results.","PeriodicalId":364463,"journal":{"name":"2022 International Conference on Emerging Technologies in Electronics, Computing and Communication (ICETECC)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127202084","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-07DOI: 10.1109/ICETECC56662.2022.10069706
Ahmad Bilal, Syed Jahania Shah, Muhammad A. Khan, Manaal Khan, Arwa Hasnain Bharmal, T. Mumtaz
The Internet of Things (IoT) refers to a telecommunication network that allows various objects to connect and share information via the Internet. It allows for machine-to-machine interaction and collaboration that does not necessitate human involvement. Its use has ushered in a new era of everything smart, including smart residences, smart cities, smart buildings, smart agriculture, and more. However, the security and privacy considerations associated with IoT pose significant hurdles. This paper provides a descriptive examination of the layered architecture of the IoT, as well as ways for overcoming security vulnerabilities using already established methods. Furthermore, based on the literature analysis, a more safe layered design that can be readily modified to improve cybersecurity problems has been proposed.
{"title":"Security Threats and Research Challenges of IoT - A Review","authors":"Ahmad Bilal, Syed Jahania Shah, Muhammad A. Khan, Manaal Khan, Arwa Hasnain Bharmal, T. Mumtaz","doi":"10.1109/ICETECC56662.2022.10069706","DOIUrl":"https://doi.org/10.1109/ICETECC56662.2022.10069706","url":null,"abstract":"The Internet of Things (IoT) refers to a telecommunication network that allows various objects to connect and share information via the Internet. It allows for machine-to-machine interaction and collaboration that does not necessitate human involvement. Its use has ushered in a new era of everything smart, including smart residences, smart cities, smart buildings, smart agriculture, and more. However, the security and privacy considerations associated with IoT pose significant hurdles. This paper provides a descriptive examination of the layered architecture of the IoT, as well as ways for overcoming security vulnerabilities using already established methods. Furthermore, based on the literature analysis, a more safe layered design that can be readily modified to improve cybersecurity problems has been proposed.","PeriodicalId":364463,"journal":{"name":"2022 International Conference on Emerging Technologies in Electronics, Computing and Communication (ICETECC)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129864940","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-07DOI: 10.1109/ICETECC56662.2022.10069212
Jamsher Bhanbhro, Shahnawaz Talpur, Asif Aziz Memon
Speech Emotion Recognition (SER) has been essential to Human-Computer Interaction (HCI) and other complex speech processing systems over the past decade. Due to the emotive differences between different speakers, SER is a complex and challenging process. The features retrieved from speech signals are crucial to SER systems’ performance. It is still challenging to develop efficient feature extracting and classification models. This study suggested hybrid deep learning models for accurately extracting crucial features and enhancing predictions with higher probabilities. Initially, the Mel spectrogram’s temporal features are trained using a combination of stacked Convolutional Neural Networks (CNN) & Long-term short memory (LSTM). The said model performs well. For enhancing the speech, samples are initially preprocessed using data improvement and dataset balancing techniques. The RAVDNESS dataset is used in this study which contains 1440 samples of audio in North American English accent. The strength of the CNN algorithm is used for obtaining spatial features and sequence encoding conversion, which generates accuracy above 93.9% for the model on mentioned data set when classifying emotions into one of eight categories. The model is generalized using Additive white Gaussian noise (AWGN) and Dropout techniques.
{"title":"Speech Emotion Recognition Using Deep Learning Hybrid Models","authors":"Jamsher Bhanbhro, Shahnawaz Talpur, Asif Aziz Memon","doi":"10.1109/ICETECC56662.2022.10069212","DOIUrl":"https://doi.org/10.1109/ICETECC56662.2022.10069212","url":null,"abstract":"Speech Emotion Recognition (SER) has been essential to Human-Computer Interaction (HCI) and other complex speech processing systems over the past decade. Due to the emotive differences between different speakers, SER is a complex and challenging process. The features retrieved from speech signals are crucial to SER systems’ performance. It is still challenging to develop efficient feature extracting and classification models. This study suggested hybrid deep learning models for accurately extracting crucial features and enhancing predictions with higher probabilities. Initially, the Mel spectrogram’s temporal features are trained using a combination of stacked Convolutional Neural Networks (CNN) & Long-term short memory (LSTM). The said model performs well. For enhancing the speech, samples are initially preprocessed using data improvement and dataset balancing techniques. The RAVDNESS dataset is used in this study which contains 1440 samples of audio in North American English accent. The strength of the CNN algorithm is used for obtaining spatial features and sequence encoding conversion, which generates accuracy above 93.9% for the model on mentioned data set when classifying emotions into one of eight categories. The model is generalized using Additive white Gaussian noise (AWGN) and Dropout techniques.","PeriodicalId":364463,"journal":{"name":"2022 International Conference on Emerging Technologies in Electronics, Computing and Communication (ICETECC)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129545233","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-17DOI: 10.1109/ICETECC56662.2022.10069052
Savera Sarwar, Muhammad Turab, Danish Channa, Aisha Chandio, M. Sohu, Vikram Kumar
One of the most important senses in human life is vision, without it one’s life is totally filled with darkness. According to WHO globally millions of people are visually impaired estimated there are 285 million, of whom some millions are blind. Unfortunately, there are around 2.4 million people are blind in our beloved country Pakistan. Human are a crucial part of society and the blind community is a main part of society. The technologies are grown so far to make the life of humans easier more comfortable and more reliable for. However, this disability of the blind community would reduce their chance of using such innovative products. Therefore, the visually impaired community believe that they are burden to other societies and they do not capture in normal activities separates the blind people from society and because of this believe did not participate in the normally tasks of society. The visual impair people mainly face most of the problems in this real-time The aim of this work is to turn the real time world into an audio world by telling blind person about the objects in their way and can read printed text. This will enable blind persons to identify the things and read the text without any external help just by using the object detection and reading system in real time. Objective of this work: i) Object detection ii) Read printed text, using state-of-the-art (SOTA) technology.
{"title":"Advanced Audio Aid for Blind People","authors":"Savera Sarwar, Muhammad Turab, Danish Channa, Aisha Chandio, M. Sohu, Vikram Kumar","doi":"10.1109/ICETECC56662.2022.10069052","DOIUrl":"https://doi.org/10.1109/ICETECC56662.2022.10069052","url":null,"abstract":"One of the most important senses in human life is vision, without it one’s life is totally filled with darkness. According to WHO globally millions of people are visually impaired estimated there are 285 million, of whom some millions are blind. Unfortunately, there are around 2.4 million people are blind in our beloved country Pakistan. Human are a crucial part of society and the blind community is a main part of society. The technologies are grown so far to make the life of humans easier more comfortable and more reliable for. However, this disability of the blind community would reduce their chance of using such innovative products. Therefore, the visually impaired community believe that they are burden to other societies and they do not capture in normal activities separates the blind people from society and because of this believe did not participate in the normally tasks of society. The visual impair people mainly face most of the problems in this real-time The aim of this work is to turn the real time world into an audio world by telling blind person about the objects in their way and can read printed text. This will enable blind persons to identify the things and read the text without any external help just by using the object detection and reading system in real time. Objective of this work: i) Object detection ii) Read printed text, using state-of-the-art (SOTA) technology.","PeriodicalId":364463,"journal":{"name":"2022 International Conference on Emerging Technologies in Electronics, Computing and Communication (ICETECC)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130230684","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-17DOI: 10.1109/ICETECC56662.2022.10069527
Wisal Khan, Muhammad Turab, Waqas Ahmad, Syed Hasnat Ahmad, Kelash Kumar, Bin Luo
Data dimension reduction (DDR) is all about mapping data from high dimensions to low dimensions, various techniques of DDR are being used for image dimension reduction like Random Projections, Principal Component Analysis (PCA), the Variance approach, LSA-Transform, the Combined and Direct approaches, and the New Random Approach. Auto-encoders (AE) are used to learn end-to-end mapping. In this paper, we demonstrate that pre-processing not only speeds up the algorithms but also improves accuracy in both supervised and unsupervised learning. In pre-processing of DDR, first PCA based DDR is used for supervised learning, then we explore AE based DDR for unsupervised learning. In PCA based DDR, we first compare supervised learning algorithms accuracy and time before and after applying PCA. Similarly, in AE based DDR, we compare unsupervised learning algorithm accuracy and time before and after AE representation learning. Supervised learning algorithms including support-vector machines (SVM), Decision Tree with GINI index, Decision Tree with entropy and Stochastic Gradient Descent classifier (SGDC) and unsupervised learning algorithm including K-means clustering, are used for classification purpose. We used two datasets MNIST and FashionMNIST Our experiment shows that there is massive improvement in accuracy and time reduction after pre-processing in both supervised and unsupervised learning.
{"title":"Data Dimension Reduction makes ML Algorithms efficient","authors":"Wisal Khan, Muhammad Turab, Waqas Ahmad, Syed Hasnat Ahmad, Kelash Kumar, Bin Luo","doi":"10.1109/ICETECC56662.2022.10069527","DOIUrl":"https://doi.org/10.1109/ICETECC56662.2022.10069527","url":null,"abstract":"Data dimension reduction (DDR) is all about mapping data from high dimensions to low dimensions, various techniques of DDR are being used for image dimension reduction like Random Projections, Principal Component Analysis (PCA), the Variance approach, LSA-Transform, the Combined and Direct approaches, and the New Random Approach. Auto-encoders (AE) are used to learn end-to-end mapping. In this paper, we demonstrate that pre-processing not only speeds up the algorithms but also improves accuracy in both supervised and unsupervised learning. In pre-processing of DDR, first PCA based DDR is used for supervised learning, then we explore AE based DDR for unsupervised learning. In PCA based DDR, we first compare supervised learning algorithms accuracy and time before and after applying PCA. Similarly, in AE based DDR, we compare unsupervised learning algorithm accuracy and time before and after AE representation learning. Supervised learning algorithms including support-vector machines (SVM), Decision Tree with GINI index, Decision Tree with entropy and Stochastic Gradient Descent classifier (SGDC) and unsupervised learning algorithm including K-means clustering, are used for classification purpose. We used two datasets MNIST and FashionMNIST Our experiment shows that there is massive improvement in accuracy and time reduction after pre-processing in both supervised and unsupervised learning.","PeriodicalId":364463,"journal":{"name":"2022 International Conference on Emerging Technologies in Electronics, Computing and Communication (ICETECC)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121927260","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}