Pub Date : 2021-01-04DOI: 10.1109/IMCOM51814.2021.9377403
Ritu Chauhan, Eiad Yafi
Bigdata in healthcare has manifested as well as benefited healthcare practioners and scientists around the globe to detect hidden patterns for future clinical decision making. The major complexity faced in real world application domain is the volume of Electronic Health Records (EHR) which has gathered due to high end IT based technology which has boomed in past century for early detection of disease. The traditional technology tools adopted were incapable to discover hidden patterns due to its computational requirements. So, Big data has its generosity need in healthcare intervene technology due to diverse nature of data and accelerated speed of data that needs to processed for better diagnostic interventions. This study has been conducted using predictive data analytics on big data for discovery of knowledge for future decision making. The study consists of information about 3,56,507 patients from 1982–2010. Data curation has been done by organizing under various categories including Age, Year (1982–2010), Incidence Counts (1982–2010, all age groups and both genders), and Mortality Counts (1982–2010, all age groups). The results represents invariable patterns which can be utilized for future predictive modelling.
{"title":"Big Data Analytics for Prediction Modelling in Healthcare Databases","authors":"Ritu Chauhan, Eiad Yafi","doi":"10.1109/IMCOM51814.2021.9377403","DOIUrl":"https://doi.org/10.1109/IMCOM51814.2021.9377403","url":null,"abstract":"Bigdata in healthcare has manifested as well as benefited healthcare practioners and scientists around the globe to detect hidden patterns for future clinical decision making. The major complexity faced in real world application domain is the volume of Electronic Health Records (EHR) which has gathered due to high end IT based technology which has boomed in past century for early detection of disease. The traditional technology tools adopted were incapable to discover hidden patterns due to its computational requirements. So, Big data has its generosity need in healthcare intervene technology due to diverse nature of data and accelerated speed of data that needs to processed for better diagnostic interventions. This study has been conducted using predictive data analytics on big data for discovery of knowledge for future decision making. The study consists of information about 3,56,507 patients from 1982–2010. Data curation has been done by organizing under various categories including Age, Year (1982–2010), Incidence Counts (1982–2010, all age groups and both genders), and Mortality Counts (1982–2010, all age groups). The results represents invariable patterns which can be utilized for future predictive modelling.","PeriodicalId":275121,"journal":{"name":"2021 15th International Conference on Ubiquitous Information Management and Communication (IMCOM)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134503288","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 : 2021-01-04DOI: 10.1109/IMCOM51814.2021.9377416
Kyu Min Yoo, R. Kil, H. Youn
This paper presents a new method of predicting the values of time series using recursive update Gaussian Kernel Function Networks. First, the input structure of time series prediction model is determined by the phase space analysis of time series. Then, the one step time series prediction model is trained using the Gaussian kernel function network. In the case of multiple step time series prediction, the estimated value is used along with previous input data to make a prediction model for the right next prediction step and the same process is recursively updated until it reaches the desired prediction step. In this model, the prediction model is trained in such a way that the accumulated error due to the recursive prediction method is reduced as much as possible. For the demonstration of the proposed method, the time series prediction of Kosdaq (one of the Korean composite index) data was performed. As a result, the proposed model outperforms other prediction models such as a simple recursive prediction model, direct prediction model and also other widely used regression methods, such as support vector machines and k-nearest neighbors.
{"title":"Time Series Prediction Based on Recursive Update Gaussian Kernel Function Networks","authors":"Kyu Min Yoo, R. Kil, H. Youn","doi":"10.1109/IMCOM51814.2021.9377416","DOIUrl":"https://doi.org/10.1109/IMCOM51814.2021.9377416","url":null,"abstract":"This paper presents a new method of predicting the values of time series using recursive update Gaussian Kernel Function Networks. First, the input structure of time series prediction model is determined by the phase space analysis of time series. Then, the one step time series prediction model is trained using the Gaussian kernel function network. In the case of multiple step time series prediction, the estimated value is used along with previous input data to make a prediction model for the right next prediction step and the same process is recursively updated until it reaches the desired prediction step. In this model, the prediction model is trained in such a way that the accumulated error due to the recursive prediction method is reduced as much as possible. For the demonstration of the proposed method, the time series prediction of Kosdaq (one of the Korean composite index) data was performed. As a result, the proposed model outperforms other prediction models such as a simple recursive prediction model, direct prediction model and also other widely used regression methods, such as support vector machines and k-nearest neighbors.","PeriodicalId":275121,"journal":{"name":"2021 15th International Conference on Ubiquitous Information Management and Communication (IMCOM)","volume":"104 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121279876","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 : 2021-01-04DOI: 10.1109/IMCOM51814.2021.9377382
Stevanus Wisnu Wiiava, I. Handoko
Twitter becomes one of the most adopted social media platforms globally, and Indonesia is a country with a rapid growth of Twitter user number in recent years. This paper discusses about the examination of conversation network of Twitter hashtag related to Covid-19 in Indonesia by using a network perspective. During this pandemic situation, Twitter has been increasingly adopted as a medium of conversational interaction amongst people to express their opinion and feeling about the situation, or share information, among others. At the same time, the Indonesian Government has established an official hashtag (#) to coordinate and organize conversations related to a specific topic of Covid-19, namely #BersatuLawanCovid19. In this way, the Government would be able to reach the public interest due to the capability of the hashtag to become a trending topic. This study examines how the Twitter conversations emerged and developed within the Twitter community by using Social Network Analysis approach. We have collected 793 Twitter users and 4441 Twitter chats from the hashtag #BersatuLawanCovid19. We then visualized the relationship network and examined the community using Social Network Analysis metrics with NodeXL. This study found that there is no a mutual engagement amongst the community members in terms of conversational practices. Interestingly, although some members of the community received a high number of engagement efforts from others, they do not actively respond to the initiatives. This suggests that the official account of government who is in-charge of managing the conversation need to enhance their communication strategy to improve the conversation within the community.
{"title":"Examining a Covid-19 Twitter Hashtag Conversation in Indonesia: A Social Network Analysis Approach","authors":"Stevanus Wisnu Wiiava, I. Handoko","doi":"10.1109/IMCOM51814.2021.9377382","DOIUrl":"https://doi.org/10.1109/IMCOM51814.2021.9377382","url":null,"abstract":"Twitter becomes one of the most adopted social media platforms globally, and Indonesia is a country with a rapid growth of Twitter user number in recent years. This paper discusses about the examination of conversation network of Twitter hashtag related to Covid-19 in Indonesia by using a network perspective. During this pandemic situation, Twitter has been increasingly adopted as a medium of conversational interaction amongst people to express their opinion and feeling about the situation, or share information, among others. At the same time, the Indonesian Government has established an official hashtag (#) to coordinate and organize conversations related to a specific topic of Covid-19, namely #BersatuLawanCovid19. In this way, the Government would be able to reach the public interest due to the capability of the hashtag to become a trending topic. This study examines how the Twitter conversations emerged and developed within the Twitter community by using Social Network Analysis approach. We have collected 793 Twitter users and 4441 Twitter chats from the hashtag #BersatuLawanCovid19. We then visualized the relationship network and examined the community using Social Network Analysis metrics with NodeXL. This study found that there is no a mutual engagement amongst the community members in terms of conversational practices. Interestingly, although some members of the community received a high number of engagement efforts from others, they do not actively respond to the initiatives. This suggests that the official account of government who is in-charge of managing the conversation need to enhance their communication strategy to improve the conversation within the community.","PeriodicalId":275121,"journal":{"name":"2021 15th International Conference on Ubiquitous Information Management and Communication (IMCOM)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126849627","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 : 2021-01-04DOI: 10.1109/IMCOM51814.2021.9377419
Min Wei, Caiqin Li, Xu Yang
With the development of industry, the consumption of industrial electricity is increasing, reducing the cost of electricity has become an urgent problem to be solved. Meanwhile, remote monitoring of connected devices and the intelligence pushed to the edges of the monitoring devices becomes critical in industrial IoT (IIoT). How to design the energy management mechanism that can respond to the change of electricity price in time is the main problem we are facing at present. This paper proposes an energy management architecture based on edge computing for industrial facility, which introduces edge computing into the factory energy management scenes. Under this architecture, an energy management mechanism based on edge computing is proposed. Finally, the proposed mechanism is tested, and the test shows that the mechanism can reduce the electricity cost of the factory.
{"title":"An Energy Management System with Edge Computing for Industrial Facility","authors":"Min Wei, Caiqin Li, Xu Yang","doi":"10.1109/IMCOM51814.2021.9377419","DOIUrl":"https://doi.org/10.1109/IMCOM51814.2021.9377419","url":null,"abstract":"With the development of industry, the consumption of industrial electricity is increasing, reducing the cost of electricity has become an urgent problem to be solved. Meanwhile, remote monitoring of connected devices and the intelligence pushed to the edges of the monitoring devices becomes critical in industrial IoT (IIoT). How to design the energy management mechanism that can respond to the change of electricity price in time is the main problem we are facing at present. This paper proposes an energy management architecture based on edge computing for industrial facility, which introduces edge computing into the factory energy management scenes. Under this architecture, an energy management mechanism based on edge computing is proposed. Finally, the proposed mechanism is tested, and the test shows that the mechanism can reduce the electricity cost of the factory.","PeriodicalId":275121,"journal":{"name":"2021 15th International Conference on Ubiquitous Information Management and Communication (IMCOM)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114871991","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 : 2021-01-04DOI: 10.1109/IMCOM51814.2021.9377396
E. Nepolo, G. Lusilao-Zodi
Due to the exponential growth of cloud computing, data centres have become the pivot for supporting the core infrastructure that propels the cloud computing evolution. Data centres are repositories that house different networking devices that are connected together using communication links to collect, store, process and disseminate data. Data centres prioritise high data availability amongst others. However, data availability is challenged by the unpredictable nature of the network environment, which presents enormous challenges in designing routing protocols that are agile enough to adjust to sudden changes in the network's available capacity. To provide seamless services to users, most modern data centres use Fat-Tree as the de-facto topology due to its multipath diversity, and the Equal-Cost Multi-Path protocol (ECMP) as the primary routing protocol to route data towards alternative paths of equal cost when the primary path is over-utilised. However, the weighted algorithm used to achieve this is inefficient, as its assigns traffic to links based on their link capacities without taking into consideration the capacity already in use on that link. In this paper, we propose the Predictive Equal-Cost Multi-Path protocol that enhances ECMP's weighted load-balancing algorithm by making forwarding decisions based on predicted congestion outlooks. The proposed protocol uses packets dropped to compute the bandwidth utilisation of links and uses the computed figures to identify the least congested links, which is then used to build forwarding tables. The protocol was tested in a Fat-Tree enabled data centre where it proved to perform better when compared to the ECMP weighted algorithm.
{"title":"A Predictive ECMP Routing Protocol for Fat-Tree Enabled Data Centre Networks","authors":"E. Nepolo, G. Lusilao-Zodi","doi":"10.1109/IMCOM51814.2021.9377396","DOIUrl":"https://doi.org/10.1109/IMCOM51814.2021.9377396","url":null,"abstract":"Due to the exponential growth of cloud computing, data centres have become the pivot for supporting the core infrastructure that propels the cloud computing evolution. Data centres are repositories that house different networking devices that are connected together using communication links to collect, store, process and disseminate data. Data centres prioritise high data availability amongst others. However, data availability is challenged by the unpredictable nature of the network environment, which presents enormous challenges in designing routing protocols that are agile enough to adjust to sudden changes in the network's available capacity. To provide seamless services to users, most modern data centres use Fat-Tree as the de-facto topology due to its multipath diversity, and the Equal-Cost Multi-Path protocol (ECMP) as the primary routing protocol to route data towards alternative paths of equal cost when the primary path is over-utilised. However, the weighted algorithm used to achieve this is inefficient, as its assigns traffic to links based on their link capacities without taking into consideration the capacity already in use on that link. In this paper, we propose the Predictive Equal-Cost Multi-Path protocol that enhances ECMP's weighted load-balancing algorithm by making forwarding decisions based on predicted congestion outlooks. The proposed protocol uses packets dropped to compute the bandwidth utilisation of links and uses the computed figures to identify the least congested links, which is then used to build forwarding tables. The protocol was tested in a Fat-Tree enabled data centre where it proved to perform better when compared to the ECMP weighted algorithm.","PeriodicalId":275121,"journal":{"name":"2021 15th International Conference on Ubiquitous Information Management and Communication (IMCOM)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129869614","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}
We present a chatbot system to offer medical consultation services to patients anytime, anywhere. Our chatbot deals with ophthalmologic diseases, currently focusing on macular degeneration. We built the system components and created QA datasets, working closely with an ophthalmologist to obtain and verify medical data.
{"title":"Developing a Ophthalmic Chatbot System","authors":"Jung-Hoon Lee, Min-Su Jeong, Jin-Uk Cho, Hyun-Kyu Jeon, Jong-Hyeok Park, Kyoung-Deok Shin, Su-Jeong Song, Yun-Gyung Cheong","doi":"10.1109/IMCOM51814.2021.9377398","DOIUrl":"https://doi.org/10.1109/IMCOM51814.2021.9377398","url":null,"abstract":"We present a chatbot system to offer medical consultation services to patients anytime, anywhere. Our chatbot deals with ophthalmologic diseases, currently focusing on macular degeneration. We built the system components and created QA datasets, working closely with an ophthalmologist to obtain and verify medical data.","PeriodicalId":275121,"journal":{"name":"2021 15th International Conference on Ubiquitous Information Management and Communication (IMCOM)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128400430","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 : 2021-01-04DOI: 10.1109/IMCOM51814.2021.9377428
Yunseon Jang, C. Son, Hyunseung Choo
In bad weather, artifacts such as rain streaks degrade the image quality. In addition, artifacts in the damaged image obstruct human vision and adversely affect the accuracy of object detection. Hence, single image rain removal is an important issue to improve image quality. However, state-of-the-art methods have limitation that require a lot of training data. This paper proposes a lightweight Deep Extraction Network (DEN), which performs well on image de-raining even with a small training dataset. Particularly, we design a novel Light Residual Block (LRB), which is connected in five cascading layers for extracting a deep local feature. Furthermore, DEN deploys a residual learning for training only artifacts. The experimental results on synthetic and real-world rainy image demonstrate the effectiveness in terms of visual and quantitative performance.
{"title":"Lightweight Deep Extraction Networks for Single Image De-raining","authors":"Yunseon Jang, C. Son, Hyunseung Choo","doi":"10.1109/IMCOM51814.2021.9377428","DOIUrl":"https://doi.org/10.1109/IMCOM51814.2021.9377428","url":null,"abstract":"In bad weather, artifacts such as rain streaks degrade the image quality. In addition, artifacts in the damaged image obstruct human vision and adversely affect the accuracy of object detection. Hence, single image rain removal is an important issue to improve image quality. However, state-of-the-art methods have limitation that require a lot of training data. This paper proposes a lightweight Deep Extraction Network (DEN), which performs well on image de-raining even with a small training dataset. Particularly, we design a novel Light Residual Block (LRB), which is connected in five cascading layers for extracting a deep local feature. Furthermore, DEN deploys a residual learning for training only artifacts. The experimental results on synthetic and real-world rainy image demonstrate the effectiveness in terms of visual and quantitative performance.","PeriodicalId":275121,"journal":{"name":"2021 15th International Conference on Ubiquitous Information Management and Communication (IMCOM)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130094386","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 : 2021-01-04DOI: 10.1109/IMCOM51814.2021.9377374
SeongGyo Seo, J. Jeon
Infrared cameras require constant nonuniformity correction because image nonuniformity occurs with environmental changes. In this paper, we propose a nonuniformity correction algorithm using feature pattern matching that can correct nonuniformities in real time. The proposed algorithm consists of motion estimation and nonuniformity correction steps. The motion estimation algorithm consists of feature extraction, feature point simplification, and feature point pattern generation steps and is proposed to calculate the amount of motion between frames in real time using a field programmable gate array. The experimental results confirm that the proposed method is robust against ghost phenomenon, compared to a statistics-based nonuniformity correction, and improves the real-time performance while providing the same performance as the existing interframe registration-based nonuniformity correction algorithm.
{"title":"Real-time scene-based nonuniformity correction using feature pattern matching","authors":"SeongGyo Seo, J. Jeon","doi":"10.1109/IMCOM51814.2021.9377374","DOIUrl":"https://doi.org/10.1109/IMCOM51814.2021.9377374","url":null,"abstract":"Infrared cameras require constant nonuniformity correction because image nonuniformity occurs with environmental changes. In this paper, we propose a nonuniformity correction algorithm using feature pattern matching that can correct nonuniformities in real time. The proposed algorithm consists of motion estimation and nonuniformity correction steps. The motion estimation algorithm consists of feature extraction, feature point simplification, and feature point pattern generation steps and is proposed to calculate the amount of motion between frames in real time using a field programmable gate array. The experimental results confirm that the proposed method is robust against ghost phenomenon, compared to a statistics-based nonuniformity correction, and improves the real-time performance while providing the same performance as the existing interframe registration-based nonuniformity correction algorithm.","PeriodicalId":275121,"journal":{"name":"2021 15th International Conference on Ubiquitous Information Management and Communication (IMCOM)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124337555","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 : 2021-01-04DOI: 10.1109/IMCOM51814.2021.9377354
Jumpei Yamada, D. Kitayama
In recent years, due to the widespread use of the Internet, the number of opportunities to search the Web using search engines has been increasing. In conventional search engines, information retrieval is achieved by repeatedly entering a query and selecting and browsing each page in the search engine result pages (SERPs). The search engines present titles, snippets, and other information to help users select suitable Web pages. However, there are cases in which people view Web pages one by one due to lack of prior knowledge or failure of search strategies. To solve this problem, we present keywords from unvisited results in the SERPs, so that users can predict the content of the Web pages. We propose two kinds of feature words as extended snippets to be presented in each search result: a content word to indicate the central content of a Web page and known-topic and unknown-topic words to indicate the degree of knowledge that one would gain by browsing the Web page. The extraction of those is based on the clustering of words in snippet sentences using the distributed representation of the words and the clustering of words in the visited pages, respectively. We investigated the impact of the proposed extended snippet on user search behavior. The experimental findings indicate that our method was useful in certain types of search, as it decreased the time necessary to complete the search. Furthermore, the participants' ratings of the extended snippets were favorable, especially those of the unknown-topic words.
{"title":"The Analysis of Web Search Snippets Displaying User's Knowledge","authors":"Jumpei Yamada, D. Kitayama","doi":"10.1109/IMCOM51814.2021.9377354","DOIUrl":"https://doi.org/10.1109/IMCOM51814.2021.9377354","url":null,"abstract":"In recent years, due to the widespread use of the Internet, the number of opportunities to search the Web using search engines has been increasing. In conventional search engines, information retrieval is achieved by repeatedly entering a query and selecting and browsing each page in the search engine result pages (SERPs). The search engines present titles, snippets, and other information to help users select suitable Web pages. However, there are cases in which people view Web pages one by one due to lack of prior knowledge or failure of search strategies. To solve this problem, we present keywords from unvisited results in the SERPs, so that users can predict the content of the Web pages. We propose two kinds of feature words as extended snippets to be presented in each search result: a content word to indicate the central content of a Web page and known-topic and unknown-topic words to indicate the degree of knowledge that one would gain by browsing the Web page. The extraction of those is based on the clustering of words in snippet sentences using the distributed representation of the words and the clustering of words in the visited pages, respectively. We investigated the impact of the proposed extended snippet on user search behavior. The experimental findings indicate that our method was useful in certain types of search, as it decreased the time necessary to complete the search. Furthermore, the participants' ratings of the extended snippets were favorable, especially those of the unknown-topic words.","PeriodicalId":275121,"journal":{"name":"2021 15th International Conference on Ubiquitous Information Management and Communication (IMCOM)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114813043","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 : 2021-01-04DOI: 10.1109/IMCOM51814.2021.9377361
V. Ponnaganti, M. Moh, Teng-Sheng Moh
This project evaluates the feasibility of utilizing popular Convolutional Neural Networks (CNNs) to detect objects present in LiDAR (Light Detection And Ranging) data, and the resulting neural network's performance. This work aims to further existing experimentation using raw LiDAR data that is analyzed and represented in a two-dimensional frame. Using this method, hundreds of frames were generated to create a dataset that was used for neural network training and validation on an existing CNN architecture. The LiDAR dataset was used to train YOLOv3, a popular CNN model, to detect vehicles. This research aims to test a smaller version of the network, YOLOv3-tiny, to measure the change in accuracy between using YOLOv3 and YOLOv3-tiny on the LiDAR dataset. The results are then compared to the loss typically found when going from YOLOv3 to YOLOV3-tiny on camera-based images. In prior experimentation, a preprocessing method was also introduced to attempt to isolate target objects in the frame. The method will be evaluated in this paper to measure its effect on the final accuracy metric of the network. Lastly, the runtime performance of these networks will be evaluated on two embedded platforms to understand if the frame rate that the networks perform on is usable for real-world applications, based on the frame rate the sensor is capable of outputting and the inference speed of the network on the embedded platforms.
{"title":"Utilizing CNNs for Object Detection with LiDAR Data for Autonomous Driving","authors":"V. Ponnaganti, M. Moh, Teng-Sheng Moh","doi":"10.1109/IMCOM51814.2021.9377361","DOIUrl":"https://doi.org/10.1109/IMCOM51814.2021.9377361","url":null,"abstract":"This project evaluates the feasibility of utilizing popular Convolutional Neural Networks (CNNs) to detect objects present in LiDAR (Light Detection And Ranging) data, and the resulting neural network's performance. This work aims to further existing experimentation using raw LiDAR data that is analyzed and represented in a two-dimensional frame. Using this method, hundreds of frames were generated to create a dataset that was used for neural network training and validation on an existing CNN architecture. The LiDAR dataset was used to train YOLOv3, a popular CNN model, to detect vehicles. This research aims to test a smaller version of the network, YOLOv3-tiny, to measure the change in accuracy between using YOLOv3 and YOLOv3-tiny on the LiDAR dataset. The results are then compared to the loss typically found when going from YOLOv3 to YOLOV3-tiny on camera-based images. In prior experimentation, a preprocessing method was also introduced to attempt to isolate target objects in the frame. The method will be evaluated in this paper to measure its effect on the final accuracy metric of the network. Lastly, the runtime performance of these networks will be evaluated on two embedded platforms to understand if the frame rate that the networks perform on is usable for real-world applications, based on the frame rate the sensor is capable of outputting and the inference speed of the network on the embedded platforms.","PeriodicalId":275121,"journal":{"name":"2021 15th International Conference on Ubiquitous Information Management and Communication (IMCOM)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126396156","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：