In the proposed model a picture-to-portray translation approach has been displayed that has consequences in colorful and sensible portrayal. The version can manipulate the fashion of various artworks. The version offers such a creative manufacturing method in a vectored environment. It additionally affords a chain of bodily applicable stroke parameters that may be used for rendering. Previous picture-to-picture translation structures have formulated the interpretation as a pixel-smart prediction. This inventive version builds a singular neural renderer that mimics the conduct of a vector renderer. Because an ordinary vector image isn't distinguishable, it defines the stroke prognosis as a factor in exploration of a method that optimizes the homology between the center and the drawing result. On parameter searching, the perception located is the zero-gradient problem. The version proposes an answer from the angle of most useful transportation. Four special strategies have additionally been compared. Metrics like SSIM, RMSE, and PSNR were used to evaluate the fineness and similarity amongst images. The layout generated via means of this research seems to be effective, and consistent with managed testing.
{"title":"Stylized NFT Progressive Neural Paintings using Brush Stroke prediction","authors":"P. Ghadekar, Prapti Maheshwari, Raj Shah, Anish Shaha, Vaishnav Sonawane, Vaibhavi Shetty","doi":"10.1109/ASSIC55218.2022.10088366","DOIUrl":"https://doi.org/10.1109/ASSIC55218.2022.10088366","url":null,"abstract":"In the proposed model a picture-to-portray translation approach has been displayed that has consequences in colorful and sensible portrayal. The version can manipulate the fashion of various artworks. The version offers such a creative manufacturing method in a vectored environment. It additionally affords a chain of bodily applicable stroke parameters that may be used for rendering. Previous picture-to-picture translation structures have formulated the interpretation as a pixel-smart prediction. This inventive version builds a singular neural renderer that mimics the conduct of a vector renderer. Because an ordinary vector image isn't distinguishable, it defines the stroke prognosis as a factor in exploration of a method that optimizes the homology between the center and the drawing result. On parameter searching, the perception located is the zero-gradient problem. The version proposes an answer from the angle of most useful transportation. Four special strategies have additionally been compared. Metrics like SSIM, RMSE, and PSNR were used to evaluate the fineness and similarity amongst images. The layout generated via means of this research seems to be effective, and consistent with managed testing.","PeriodicalId":441406,"journal":{"name":"2022 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131890497","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 : 2022-11-19DOI: 10.1109/ASSIC55218.2022.10088365
G. S. Kumar, B. Ankayarkanni
Our body's natural safeguarding in an intensified psycho-physical situation is stress. Stress inducing factors are stressors, which may affect a person's physical or mental state due to an extended exposure to various stressors. So, an efficient stress monitoring mechanism is highly essential in this busy daily environment. Stress can be identified as biological signals which can be psychological or physiological signal. In this paper a detailed study of various sources for identifying stress and an efficient way for classifying or detecting this stress by various machine learning and deep learning techniques has performed. Here various stressors, methodologies, outcomes, benefits, limitations for various stress detection techniques are highlighted which can serve as a guide for further investigations.
{"title":"Comparative Study on Mental Stress Detection Using Various Stressors and Classification Techniques","authors":"G. S. Kumar, B. Ankayarkanni","doi":"10.1109/ASSIC55218.2022.10088365","DOIUrl":"https://doi.org/10.1109/ASSIC55218.2022.10088365","url":null,"abstract":"Our body's natural safeguarding in an intensified psycho-physical situation is stress. Stress inducing factors are stressors, which may affect a person's physical or mental state due to an extended exposure to various stressors. So, an efficient stress monitoring mechanism is highly essential in this busy daily environment. Stress can be identified as biological signals which can be psychological or physiological signal. In this paper a detailed study of various sources for identifying stress and an efficient way for classifying or detecting this stress by various machine learning and deep learning techniques has performed. Here various stressors, methodologies, outcomes, benefits, limitations for various stress detection techniques are highlighted which can serve as a guide for further investigations.","PeriodicalId":441406,"journal":{"name":"2022 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128338741","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 : 2022-11-19DOI: 10.1109/ASSIC55218.2022.10088393
Rajan Singh, K. Rao, Raju Naik, Geetha, K. Anjali, P. Vineeth
Shopping is both a regular and a tedious process, especially in a supermarket. Though the internet has revolutionized the way of retail shopping, online shopping does not look to fully replace the brick and mortar stores. However, large crowd during peak hours and on public holidays generally led to longer billing time in the retail stores makes customers impatient, and also affects overall shopping experience. Here, the present work presents a novel and smart design of shopping trolley using RFID technology and Arduino which facilitates shoppers to scan the products on their own. Furthermore, the proposed system also provide a web interface for generating the bill and provide an automated payment interface for the customer. Subsequently, the proposed model is expected to ease off queue pressure at billing counters, and offers enhanced shopping experience to the customers. The same has been realized with the help of module ESP8266, which provides microcontrollers connection to 2.4 GHz Wi-Fi. The results show reduced billing time and enhanced customer experience. Additionally, fixed-cost per store is reduced that allows for further leveraging of existing resources and possible expansion of stores.
{"title":"Smart Trolley Using Automated Billing Interface","authors":"Rajan Singh, K. Rao, Raju Naik, Geetha, K. Anjali, P. Vineeth","doi":"10.1109/ASSIC55218.2022.10088393","DOIUrl":"https://doi.org/10.1109/ASSIC55218.2022.10088393","url":null,"abstract":"Shopping is both a regular and a tedious process, especially in a supermarket. Though the internet has revolutionized the way of retail shopping, online shopping does not look to fully replace the brick and mortar stores. However, large crowd during peak hours and on public holidays generally led to longer billing time in the retail stores makes customers impatient, and also affects overall shopping experience. Here, the present work presents a novel and smart design of shopping trolley using RFID technology and Arduino which facilitates shoppers to scan the products on their own. Furthermore, the proposed system also provide a web interface for generating the bill and provide an automated payment interface for the customer. Subsequently, the proposed model is expected to ease off queue pressure at billing counters, and offers enhanced shopping experience to the customers. The same has been realized with the help of module ESP8266, which provides microcontrollers connection to 2.4 GHz Wi-Fi. The results show reduced billing time and enhanced customer experience. Additionally, fixed-cost per store is reduced that allows for further leveraging of existing resources and possible expansion of stores.","PeriodicalId":441406,"journal":{"name":"2022 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130651930","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 : 2022-11-19DOI: 10.1109/ASSIC55218.2022.10088306
Ahmad Wael Mahmoud, Raed M. T. Abdulla, Muhammad Ehsan Rana, H. K. Tripathy
Energy Management Systems (EMS) provide information on energy usage, especially which device is consuming how much energy for monitoring and control. These EMS can be substantially improved and enhanced through the use of Internet of Things (IoT) based energy monitoring technology to save more energy. This research proposes a real-time IoT based energy management system for smart green buildings. The proposed system contains three main phases, including measuring power consumption, forecasting power consumption, and face recognition. The method of forecasting used in this research is Short-Term Load Forecasting (STLF), based on the K-Nearest Neighbor (KNN) algorithm. There are six variables from Digital Power Meter (DPM) required as reference data to train the prediction methods, including Line Current A, Line Current B, Line Current C, Line Voltage A, Line Voltage B, and Line Voltage C. The forecasted result determines the power consumption of the smart building for the next hours of the same day. The active, reactive, and apparent powers are calculated based on the forecasted result. Face recognition in a smart building can prevent unauthorized persons from entering a certain area of a smart building. The method used in face recognition is based on the Viola-Johns algorithm. The results obtained from the accuracy of the Viola-Johns classifier based on Haar features indicate that the system can perfectly detect and recognize faces with a total accuracy of 90%. The True Negative Rate (TNR), Positive Predictive Value (PPV) and False Discovery Rate (FDR) were found to be 50%, 69.4%, and 30.5%, respectively.
能源管理系统(EMS)提供有关能源使用的信息，特别是哪个设备消耗了多少能源进行监测和控制。通过使用基于物联网(IoT)的能源监测技术，可以大大改善和增强这些环境管理系统，从而节省更多的能源。本研究提出了一种基于物联网的智能绿色建筑实时能源管理系统。该系统包括三个主要阶段:测量功耗、预测功耗和人脸识别。本研究使用的预测方法是基于k -最近邻(KNN)算法的短期负荷预测(STLF)。需要DPM (Digital Power Meter)中的6个变量作为训练预测方法的参考数据，包括“线路电流A”、“线路电流B”、“线路电流C”、“线路电压A”、“线路电压B”和“线路电压C”。预测结果决定了智能建筑当天接下来几个小时的用电量。根据预测结果计算有功、无功和视在功率。智能楼宇中的人脸识别可以防止未经授权的人员进入智能楼宇的特定区域。人脸识别中使用的方法是基于Viola-Johns算法。从基于Haar特征的Viola-Johns分类器的准确率得到的结果表明，该系统可以很好地检测和识别人脸，总准确率达到90%。真实阴性率(TNR)、阳性预测值(PPV)和错误发现率(FDR)分别为50%、69.4%和30.5%。
{"title":"IoT Based Energy Management Solution for Smart Green Buildings","authors":"Ahmad Wael Mahmoud, Raed M. T. Abdulla, Muhammad Ehsan Rana, H. K. Tripathy","doi":"10.1109/ASSIC55218.2022.10088306","DOIUrl":"https://doi.org/10.1109/ASSIC55218.2022.10088306","url":null,"abstract":"Energy Management Systems (EMS) provide information on energy usage, especially which device is consuming how much energy for monitoring and control. These EMS can be substantially improved and enhanced through the use of Internet of Things (IoT) based energy monitoring technology to save more energy. This research proposes a real-time IoT based energy management system for smart green buildings. The proposed system contains three main phases, including measuring power consumption, forecasting power consumption, and face recognition. The method of forecasting used in this research is Short-Term Load Forecasting (STLF), based on the K-Nearest Neighbor (KNN) algorithm. There are six variables from Digital Power Meter (DPM) required as reference data to train the prediction methods, including Line Current A, Line Current B, Line Current C, Line Voltage A, Line Voltage B, and Line Voltage C. The forecasted result determines the power consumption of the smart building for the next hours of the same day. The active, reactive, and apparent powers are calculated based on the forecasted result. Face recognition in a smart building can prevent unauthorized persons from entering a certain area of a smart building. The method used in face recognition is based on the Viola-Johns algorithm. The results obtained from the accuracy of the Viola-Johns classifier based on Haar features indicate that the system can perfectly detect and recognize faces with a total accuracy of 90%. The True Negative Rate (TNR), Positive Predictive Value (PPV) and False Discovery Rate (FDR) were found to be 50%, 69.4%, and 30.5%, respectively.","PeriodicalId":441406,"journal":{"name":"2022 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121587850","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 : 2022-11-19DOI: 10.1109/ASSIC55218.2022.10088313
Lai Yi Heng, Muhammad Ehsan Rana, Raed M. T. Abdulla, H. K. Tripathy
A weather station uses many sensors to collect ambient data. With the help of IoT, it integrates all devices, connects them to the Internet and frames an intelligent ecosystem. For this research, a smart weather monitoring system has been proposed which detects temperature, humidity, pressure, altitude, dew point, and light level, as well as the existence of water in a specific place. All data is shown on the OLED screen and then displayed on the Ubidots website and app. If the vicinity is dark, LED will turn on automatically, and an alert sound will be generated if water is detected. The system also sends SMS based on certain conditions. Authors have used various hardware devices, including Node32 Lite, BMP280, DHT22, MH-RD, OLED, LED, LDR, piezo buzzer, etc., in providing the proof of concept.
{"title":"Design and Development Recommendations for a Smart Weather Monitoring System","authors":"Lai Yi Heng, Muhammad Ehsan Rana, Raed M. T. Abdulla, H. K. Tripathy","doi":"10.1109/ASSIC55218.2022.10088313","DOIUrl":"https://doi.org/10.1109/ASSIC55218.2022.10088313","url":null,"abstract":"A weather station uses many sensors to collect ambient data. With the help of IoT, it integrates all devices, connects them to the Internet and frames an intelligent ecosystem. For this research, a smart weather monitoring system has been proposed which detects temperature, humidity, pressure, altitude, dew point, and light level, as well as the existence of water in a specific place. All data is shown on the OLED screen and then displayed on the Ubidots website and app. If the vicinity is dark, LED will turn on automatically, and an alert sound will be generated if water is detected. The system also sends SMS based on certain conditions. Authors have used various hardware devices, including Node32 Lite, BMP280, DHT22, MH-RD, OLED, LED, LDR, piezo buzzer, etc., in providing the proof of concept.","PeriodicalId":441406,"journal":{"name":"2022 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC)","volume":"246 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114279086","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 : 2022-11-19DOI: 10.1109/ASSIC55218.2022.10088390
P. Purushotham, D. D. Priya, A. Kiran
Disaster analysis includes data on natural and man-made disasters like tsunamis and earthquakes. This article reviews machine learning techniques for pandemic and disaster management. Most nations worry about rare disasters and pandemics. Disaster and pandemic management has used IoT, object sensing, UAV, 5G and cellular networks, smartphone-based systems, and satellite-based systems. Machine learning (ML) methods can handle multidimensional, enormous volumes of data found in disaster and pandemic management and are well-suited for related tasks such as recognition and classification. Machine learning algorithms can predict disasters and help with disaster management duties including establishing crowd evacuation routes and analyzing social media posts. Machine learning algorithms also help anticipate pandemics, monitor pandemic spread, and diagnose diseases.
{"title":"Disaster Analysis Using Machine Learning","authors":"P. Purushotham, D. D. Priya, A. Kiran","doi":"10.1109/ASSIC55218.2022.10088390","DOIUrl":"https://doi.org/10.1109/ASSIC55218.2022.10088390","url":null,"abstract":"Disaster analysis includes data on natural and man-made disasters like tsunamis and earthquakes. This article reviews machine learning techniques for pandemic and disaster management. Most nations worry about rare disasters and pandemics. Disaster and pandemic management has used IoT, object sensing, UAV, 5G and cellular networks, smartphone-based systems, and satellite-based systems. Machine learning (ML) methods can handle multidimensional, enormous volumes of data found in disaster and pandemic management and are well-suited for related tasks such as recognition and classification. Machine learning algorithms can predict disasters and help with disaster management duties including establishing crowd evacuation routes and analyzing social media posts. Machine learning algorithms also help anticipate pandemics, monitor pandemic spread, and diagnose diseases.","PeriodicalId":441406,"journal":{"name":"2022 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114671454","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 : 2022-11-19DOI: 10.1109/ASSIC55218.2022.10088395
Yeo Zi Jian, Noris Ismail, Moin Nabi
The flexibility of Wi-Fi connection has made people more dependent on wireless connection, due to its flexibility and mobility which is highly suited for the current fast paced society. Wi-Fi connections is highly preferred compared to normal wired connection. Wi-Fi signals is generated by a Wireless Access Point (WAP) device, oblivious mindset towards WAP positioning often result in poor Wi-Fi connectivity. On the flip side, proper placement of WAP is crucial to ensure minimum signal wastage and maximum signal coverage in each compound. The report below highlights the research & development on a web-based application named Wireless Access Point Mapper (WAP-MAP). The proposed system functions to predict then recommend optimal WAP placement and the appropriate quantity in each 2-Dimensional floor plan, while equipped with multiple security features to ensure both functionality & security.
{"title":"Wireless Access Point Mapper (WAP-MAP): An Automated Wireless Access Point Plotting Web Application","authors":"Yeo Zi Jian, Noris Ismail, Moin Nabi","doi":"10.1109/ASSIC55218.2022.10088395","DOIUrl":"https://doi.org/10.1109/ASSIC55218.2022.10088395","url":null,"abstract":"The flexibility of Wi-Fi connection has made people more dependent on wireless connection, due to its flexibility and mobility which is highly suited for the current fast paced society. Wi-Fi connections is highly preferred compared to normal wired connection. Wi-Fi signals is generated by a Wireless Access Point (WAP) device, oblivious mindset towards WAP positioning often result in poor Wi-Fi connectivity. On the flip side, proper placement of WAP is crucial to ensure minimum signal wastage and maximum signal coverage in each compound. The report below highlights the research & development on a web-based application named Wireless Access Point Mapper (WAP-MAP). The proposed system functions to predict then recommend optimal WAP placement and the appropriate quantity in each 2-Dimensional floor plan, while equipped with multiple security features to ensure both functionality & security.","PeriodicalId":441406,"journal":{"name":"2022 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124250734","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 : 2022-11-19DOI: 10.1109/ASSIC55218.2022.10088336
E. Anupriya, N. Kumaresan, V. Suresh, S. Dhanasekaran, K. Ramprathap, P. Chinnasamy
Nowadays, a person's impact is frequently determined by the number of followers he or she has on social media. To this aim, the prevalence of false accounts is one of the most pressing issues, with the potential to disrupt a wide range of real-world and economic activity. Bot followers are dangerous to social media as these could alter perceptions of popularity and influence, which can have a ample amount of impact on every sector. As a result, new approaches must be developed to enable the detection and classification of bogus accounts. This study gives novel method for distinguishing original profiles. The method uses information gathered automatically from huge data to characterize typical patterns of fake account.
{"title":"Fraud Account Detection on Social Network using Machine Learning Techniques","authors":"E. Anupriya, N. Kumaresan, V. Suresh, S. Dhanasekaran, K. Ramprathap, P. Chinnasamy","doi":"10.1109/ASSIC55218.2022.10088336","DOIUrl":"https://doi.org/10.1109/ASSIC55218.2022.10088336","url":null,"abstract":"Nowadays, a person's impact is frequently determined by the number of followers he or she has on social media. To this aim, the prevalence of false accounts is one of the most pressing issues, with the potential to disrupt a wide range of real-world and economic activity. Bot followers are dangerous to social media as these could alter perceptions of popularity and influence, which can have a ample amount of impact on every sector. As a result, new approaches must be developed to enable the detection and classification of bogus accounts. This study gives novel method for distinguishing original profiles. The method uses information gathered automatically from huge data to characterize typical patterns of fake account.","PeriodicalId":441406,"journal":{"name":"2022 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125303058","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 : 2022-11-19DOI: 10.1109/ASSIC55218.2022.10088403
A. Kiran, P. Purushotham, D. D. Priya
Increased crime in packed events or lonely areas has made security a top priority in every industry. Computer Vision is used to find and fix anomalies. Increasing needs for security, privacy, and private property protection require video surveillance systems that can recognize and understand scene and anomalous situations. Monitoring such activities and recognizing antisocial behavior helps minimize crime and social offenses. Existing surveillance and control systems need human oversight. We're interested in detecting firearms quickly through photos and surveillance data. We recast the detection problem as decreasing false positives and solve it by building a data set guided by a deep CNN classifier and evaluating the best classification model using the region proposal approach. Our model uses Faster RCNN, YOLO.
{"title":"Weapon Detection using Artificial Intelligence and Deep Learning for Security Applications","authors":"A. Kiran, P. Purushotham, D. D. Priya","doi":"10.1109/ASSIC55218.2022.10088403","DOIUrl":"https://doi.org/10.1109/ASSIC55218.2022.10088403","url":null,"abstract":"Increased crime in packed events or lonely areas has made security a top priority in every industry. Computer Vision is used to find and fix anomalies. Increasing needs for security, privacy, and private property protection require video surveillance systems that can recognize and understand scene and anomalous situations. Monitoring such activities and recognizing antisocial behavior helps minimize crime and social offenses. Existing surveillance and control systems need human oversight. We're interested in detecting firearms quickly through photos and surveillance data. We recast the detection problem as decreasing false positives and solve it by building a data set guided by a deep CNN classifier and evaluating the best classification model using the region proposal approach. Our model uses Faster RCNN, YOLO.","PeriodicalId":441406,"journal":{"name":"2022 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127775629","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 : 2022-11-19DOI: 10.1109/ASSIC55218.2022.10088310
Noor Jehan Ashaari Muhamad, Muhammad Ehsan Rana, Vazeerudeen Abdul Hameed, H. K. Tripathy
Potholes are a regular occurrence that can cause discomfort and harm everyday road users. In recent times many studies have been done on automated pothole detection as there is a need to assess the road condition in a more affordable and timely manner. This research aims to explore the different motion-based approaches used in pothole detection. Motion sensors such as accelerometers and gyroscopes are commonly utilised to acquire movement information, and these data can be used not only to detect the presence of potholes but also have been used to classify general road conditions. It has been found that the approaches can be divided into two categories: threshold-based and machine learning. For both approaches, statistical features are extracted from the motion data and used in determining the threshold values or as inputs to train the classifier models. Further opportunities for improvement in data labelling and the need to classify pothole severity levels using a standard metric are also discussed in the paper.
{"title":"Machine Learning Combined with Thresholding - A Blended Approach to Potholes Detection","authors":"Noor Jehan Ashaari Muhamad, Muhammad Ehsan Rana, Vazeerudeen Abdul Hameed, H. K. Tripathy","doi":"10.1109/ASSIC55218.2022.10088310","DOIUrl":"https://doi.org/10.1109/ASSIC55218.2022.10088310","url":null,"abstract":"Potholes are a regular occurrence that can cause discomfort and harm everyday road users. In recent times many studies have been done on automated pothole detection as there is a need to assess the road condition in a more affordable and timely manner. This research aims to explore the different motion-based approaches used in pothole detection. Motion sensors such as accelerometers and gyroscopes are commonly utilised to acquire movement information, and these data can be used not only to detect the presence of potholes but also have been used to classify general road conditions. It has been found that the approaches can be divided into two categories: threshold-based and machine learning. For both approaches, statistical features are extracted from the motion data and used in determining the threshold values or as inputs to train the classifier models. Further opportunities for improvement in data labelling and the need to classify pothole severity levels using a standard metric are also discussed in the paper.","PeriodicalId":441406,"journal":{"name":"2022 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121666168","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：