Pub Date : 2021-11-06DOI: 10.1109/ICSET53708.2021.9612572
Alsaleh Saad, Sabiroh Md Sabri, H. Haron
This paper reports the development of a model system through which academicians can exchange and share research knowledge (RK) in higher learning institutions (HLI). The underlying research to support this development adopted a mixed-method design using an exploratory sequential approach. Data were collected from academic staff in a Malaysian public university over a period of six months using questionnaires and in-person interviews. Quantitative and qualitative analyses were used to extract and describe the study's results. The study found that the academicians share ten types of RK, including research activities, research topics, research methods, data analysis techniques, research findings, research proposals, research papers, publication procedures, research on subject areas, and research innovation. Based on the study findings, a research knowledge system model (RKSM) was developed. The study contributes to the knowledge sharing area through the identification of RK types shared in HLI, drawn from the quantitative and qualitative results. HLI administrators could benefit from the study findings and establish policies to utilize these types of knowledge.
{"title":"Research Knowledge System Model for Higher Learning Institutions","authors":"Alsaleh Saad, Sabiroh Md Sabri, H. Haron","doi":"10.1109/ICSET53708.2021.9612572","DOIUrl":"https://doi.org/10.1109/ICSET53708.2021.9612572","url":null,"abstract":"This paper reports the development of a model system through which academicians can exchange and share research knowledge (RK) in higher learning institutions (HLI). The underlying research to support this development adopted a mixed-method design using an exploratory sequential approach. Data were collected from academic staff in a Malaysian public university over a period of six months using questionnaires and in-person interviews. Quantitative and qualitative analyses were used to extract and describe the study's results. The study found that the academicians share ten types of RK, including research activities, research topics, research methods, data analysis techniques, research findings, research proposals, research papers, publication procedures, research on subject areas, and research innovation. Based on the study findings, a research knowledge system model (RKSM) was developed. The study contributes to the knowledge sharing area through the identification of RK types shared in HLI, drawn from the quantitative and qualitative results. HLI administrators could benefit from the study findings and establish policies to utilize these types of knowledge.","PeriodicalId":433197,"journal":{"name":"2021 IEEE 11th International Conference on System Engineering and Technology (ICSET)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131223596","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-06DOI: 10.1109/ICSET53708.2021.9612556
Nadher Alsafwani, Musab A. M. Ali, N. Tahir
Security is a major concern for granting protected communication among mobile nodes in an unfavorable environment. Wireless Ad Hoc can be vulnerable against attacks through malicious nodes. Thus, each node must be prepared to deal with both direct and indirect attacks. Hence, this study investigates in assessing the effect of eavesdropping attacks on Mobile Ad Hoc network systems (MANET's) using QualNet simulator. In addition, the MANET performance with eavesdropping attacks is measured. The simulation was repeated nine times on the network layer along with the data link layer of Mobile nodes in a wireless Ad Hoc network. The MANET performance is examined and what-if analysis is done to improve the mobile nodes. Results showed that the proposed method is apt for developing Mobile Ad Hoc nodes for security purposes.
{"title":"Performance Evaluation of the Mobile Ad Hoc Network (MANET) for Eavesdropping Attacks by QualN et Simulator","authors":"Nadher Alsafwani, Musab A. M. Ali, N. Tahir","doi":"10.1109/ICSET53708.2021.9612556","DOIUrl":"https://doi.org/10.1109/ICSET53708.2021.9612556","url":null,"abstract":"Security is a major concern for granting protected communication among mobile nodes in an unfavorable environment. Wireless Ad Hoc can be vulnerable against attacks through malicious nodes. Thus, each node must be prepared to deal with both direct and indirect attacks. Hence, this study investigates in assessing the effect of eavesdropping attacks on Mobile Ad Hoc network systems (MANET's) using QualNet simulator. In addition, the MANET performance with eavesdropping attacks is measured. The simulation was repeated nine times on the network layer along with the data link layer of Mobile nodes in a wireless Ad Hoc network. The MANET performance is examined and what-if analysis is done to improve the mobile nodes. Results showed that the proposed method is apt for developing Mobile Ad Hoc nodes for security purposes.","PeriodicalId":433197,"journal":{"name":"2021 IEEE 11th International Conference on System Engineering and Technology (ICSET)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130366091","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-06DOI: 10.1109/ICSET53708.2021.9612532
L. K. Haw, Nur Atiqah Jefry, Wong Kiing Ing
This paper extended the findings of the previously proposed multilevel inverter (MLI) topology to produce 51-level of AC output voltage waveform in hybrid configuration. The proposed topology is structured by merging the feature of Cascaded H-bridge Multilevel inverter (CHB-MLI) and the Diode Clamped Multilevel Inverter (DC-MLI). When compared with the previous work, this proposed topology utilized 1 additional DC voltage source plus 4 switches (i.e., making it a total of 23 components) to achieve the aforementioned outcome. With such great number of voltage level being generated, the proposed topology also meets the total harmonic distortion (THD) limit set by IEEE standard (i.e., 5%) across the selected ranges of modulation indexes (i.e., 0.3 to 1). Through the simulation conducted via Matlab/Simulink, the variations between the number of voltage level, THD, and its RMS voltage are being analyzed and thoroughly discussed. The comparison of the proposed topology against the recently presented topologies are also carried to strengthen its novelty.
{"title":"The New Hybrid Multilevel Inverter with Reduced Number of Switches","authors":"L. K. Haw, Nur Atiqah Jefry, Wong Kiing Ing","doi":"10.1109/ICSET53708.2021.9612532","DOIUrl":"https://doi.org/10.1109/ICSET53708.2021.9612532","url":null,"abstract":"This paper extended the findings of the previously proposed multilevel inverter (MLI) topology to produce 51-level of AC output voltage waveform in hybrid configuration. The proposed topology is structured by merging the feature of Cascaded H-bridge Multilevel inverter (CHB-MLI) and the Diode Clamped Multilevel Inverter (DC-MLI). When compared with the previous work, this proposed topology utilized 1 additional DC voltage source plus 4 switches (i.e., making it a total of 23 components) to achieve the aforementioned outcome. With such great number of voltage level being generated, the proposed topology also meets the total harmonic distortion (THD) limit set by IEEE standard (i.e., 5%) across the selected ranges of modulation indexes (i.e., 0.3 to 1). Through the simulation conducted via Matlab/Simulink, the variations between the number of voltage level, THD, and its RMS voltage are being analyzed and thoroughly discussed. The comparison of the proposed topology against the recently presented topologies are also carried to strengthen its novelty.","PeriodicalId":433197,"journal":{"name":"2021 IEEE 11th International Conference on System Engineering and Technology (ICSET)","volume":"142 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116586920","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-06DOI: 10.1109/ICSET53708.2021.9612557
Suit Mun Ng, H. Yazid, S. A. Rahim, N. Mustafa
Image contrast enhancement is known as one of the important techniques applied in the field of image processing. In order to improve the contrast of the captured image, different adaptive Unsharp Masking Filter (UMF) techniques were proposed by the researchers. In this paper, the main contribution is the implementation of three algorithms namely adaptive gain adjustment approach using an UMF (ASAUMF), design of UMF kernel and gain using Particle Swarm Optimization (UMFKG) and lastly, intensity and edge-based adaptive UMF (IntEdgUMF) which is denoted as Algorithm 1, 2 and 3 respectively. These algorithms were tested on the standard and biometric images like face images. This is because these adaptive UMF were mainly applied to natural scenery, but the importance of high image quality is not limited to the environment but also to the other fields such as biometric identification. Based on the results, Algorithm 1 is able to achieve the highest average PSNR values of 31.6079 dB and 35.8052 dB when applied on Set14 and LFW databases respectively. Although Algorithm 1 needs a longer running time in producing the output images, this algorithm can emphasize the details or information from the input image by enhancing the contrast of the image. Thus, Algorithm 1 can be concluded as the best adaptive UMF techniques among the three algorithms tested. For future work, the use of these adaptive UMF can be implemented onto various images, for instance gray scale images or other biometric images in order to test the effectiveness of the algorithms in different applications.
{"title":"Performance Analysis of Adaptive Unsharp Masking Filter Techniques for Image Contrast Enhancement","authors":"Suit Mun Ng, H. Yazid, S. A. Rahim, N. Mustafa","doi":"10.1109/ICSET53708.2021.9612557","DOIUrl":"https://doi.org/10.1109/ICSET53708.2021.9612557","url":null,"abstract":"Image contrast enhancement is known as one of the important techniques applied in the field of image processing. In order to improve the contrast of the captured image, different adaptive Unsharp Masking Filter (UMF) techniques were proposed by the researchers. In this paper, the main contribution is the implementation of three algorithms namely adaptive gain adjustment approach using an UMF (ASAUMF), design of UMF kernel and gain using Particle Swarm Optimization (UMFKG) and lastly, intensity and edge-based adaptive UMF (IntEdgUMF) which is denoted as Algorithm 1, 2 and 3 respectively. These algorithms were tested on the standard and biometric images like face images. This is because these adaptive UMF were mainly applied to natural scenery, but the importance of high image quality is not limited to the environment but also to the other fields such as biometric identification. Based on the results, Algorithm 1 is able to achieve the highest average PSNR values of 31.6079 dB and 35.8052 dB when applied on Set14 and LFW databases respectively. Although Algorithm 1 needs a longer running time in producing the output images, this algorithm can emphasize the details or information from the input image by enhancing the contrast of the image. Thus, Algorithm 1 can be concluded as the best adaptive UMF techniques among the three algorithms tested. For future work, the use of these adaptive UMF can be implemented onto various images, for instance gray scale images or other biometric images in order to test the effectiveness of the algorithms in different applications.","PeriodicalId":433197,"journal":{"name":"2021 IEEE 11th International Conference on System Engineering and Technology (ICSET)","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122937026","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-06DOI: 10.1109/ICSET53708.2021.9612434
A. Putra, B. Trilaksono, E. Hidayat
Hybrid Autonomous Underwater Glider (HAUG) is a vehicle used for underwater missions such as monitoring and finding new underwater resources. HAUG has good endurance and maneuverability compared to conventional Autonomous Underwater Vehicle (AUV) and Autonomous Underwater Glider (AUG). It is because HAUG has two operational modes. They are AUV and AUG's operational mode. When HAUG is in some missions, it may be faced with an obstacle that can be a threat to the HUG's safety. Therefore, HAUG should have the ability to detect and avoid obstacles. Gemini 720 im Imaging Forward Looking Sonar (FLS) is used for obstacle detection in this work. The main issue of underwater obstacle detection is noisy data received by sonar. Therefore, by designing an obstacle detection, it will overcome those issues. Frost filter and local histogram entropy are used in the sonar data processing. The processed sonar data are provided in the local sonar frame then will be used by obstacle avoidance systems. BK-product fuzzy and reactive algorithms are used for obstacle avoidance. In this paper, we added some procedures to those obstacle avoidance algorithms to handle the huge or non-complex u-shaped obstacle. Both of the obstacle detection and avoidance simulations are in ROS (Robot Operating System). The obstacle detection simulation shows that the different sizes of obstacles can be detected with average errors of approximately 0.335 meters. The obstacle avoidance simulations are in AUV's mode with no ocean current applied. The obstacle avoidance simulated in this work is with two cases. Using simulated lidar as a sensor's output and using sonar's plugin provided by Gazebo. The obstacle avoidance using simulated lidar shows that the error's value is approximately 10.12 meters, 103.62 meters, and 354.4 meters respectively. The obstacle avoidance simulation with sonar's plugin shows that the error's value is 6.55 meters.
{"title":"Hybrid Autonomous Underwater Glider (HAUG) Obstacle Detection and Avoidance","authors":"A. Putra, B. Trilaksono, E. Hidayat","doi":"10.1109/ICSET53708.2021.9612434","DOIUrl":"https://doi.org/10.1109/ICSET53708.2021.9612434","url":null,"abstract":"Hybrid Autonomous Underwater Glider (HAUG) is a vehicle used for underwater missions such as monitoring and finding new underwater resources. HAUG has good endurance and maneuverability compared to conventional Autonomous Underwater Vehicle (AUV) and Autonomous Underwater Glider (AUG). It is because HAUG has two operational modes. They are AUV and AUG's operational mode. When HAUG is in some missions, it may be faced with an obstacle that can be a threat to the HUG's safety. Therefore, HAUG should have the ability to detect and avoid obstacles. Gemini 720 im Imaging Forward Looking Sonar (FLS) is used for obstacle detection in this work. The main issue of underwater obstacle detection is noisy data received by sonar. Therefore, by designing an obstacle detection, it will overcome those issues. Frost filter and local histogram entropy are used in the sonar data processing. The processed sonar data are provided in the local sonar frame then will be used by obstacle avoidance systems. BK-product fuzzy and reactive algorithms are used for obstacle avoidance. In this paper, we added some procedures to those obstacle avoidance algorithms to handle the huge or non-complex u-shaped obstacle. Both of the obstacle detection and avoidance simulations are in ROS (Robot Operating System). The obstacle detection simulation shows that the different sizes of obstacles can be detected with average errors of approximately 0.335 meters. The obstacle avoidance simulations are in AUV's mode with no ocean current applied. The obstacle avoidance simulated in this work is with two cases. Using simulated lidar as a sensor's output and using sonar's plugin provided by Gazebo. The obstacle avoidance using simulated lidar shows that the error's value is approximately 10.12 meters, 103.62 meters, and 354.4 meters respectively. The obstacle avoidance simulation with sonar's plugin shows that the error's value is 6.55 meters.","PeriodicalId":433197,"journal":{"name":"2021 IEEE 11th International Conference on System Engineering and Technology (ICSET)","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125088732","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-06DOI: 10.1109/ICSET53708.2021.9612436
Noor Aldeen A. Khalid, Muhammad Imran Ahmad, Thulfiqar H. Mandeel, M. I. N. Isa, Raja Abdullah Raja Ahmad, Mustafa Zuhaer Nayef Al-Dabagh
This paper proposes new technique to extract the Region of Interest (ROI) of palmprint biometric image while removing the distortion between images such as translation or rotation during ROI extraction. A similarity measure known as Enhanced Correlation Coefficient (ECC) is used in the proposed approach for better ROI extraction and image alignment, which helps to evaluate and determine the distortion. The objective of image alignment approaches are to find the deformation or transformation that minimizes the incongruities between images. After applying ECC algorithm the Region of Interest (ROI) is extracted from the palmprint by using moore neighbors algorithm, on the other hand, to verify and validate the efficacy of the recommended method the PolyU palmprint dataset II was used. The results show the high accuracy is 99.8% in deriving the ROI and developing a robust ROI cropping system successfully.
{"title":"Palmprint ROI Cropping Based on Enhanced Correlation Coefficient Maximisation Algorithm","authors":"Noor Aldeen A. Khalid, Muhammad Imran Ahmad, Thulfiqar H. Mandeel, M. I. N. Isa, Raja Abdullah Raja Ahmad, Mustafa Zuhaer Nayef Al-Dabagh","doi":"10.1109/ICSET53708.2021.9612436","DOIUrl":"https://doi.org/10.1109/ICSET53708.2021.9612436","url":null,"abstract":"This paper proposes new technique to extract the Region of Interest (ROI) of palmprint biometric image while removing the distortion between images such as translation or rotation during ROI extraction. A similarity measure known as Enhanced Correlation Coefficient (ECC) is used in the proposed approach for better ROI extraction and image alignment, which helps to evaluate and determine the distortion. The objective of image alignment approaches are to find the deformation or transformation that minimizes the incongruities between images. After applying ECC algorithm the Region of Interest (ROI) is extracted from the palmprint by using moore neighbors algorithm, on the other hand, to verify and validate the efficacy of the recommended method the PolyU palmprint dataset II was used. The results show the high accuracy is 99.8% in deriving the ROI and developing a robust ROI cropping system successfully.","PeriodicalId":433197,"journal":{"name":"2021 IEEE 11th International Conference on System Engineering and Technology (ICSET)","volume":"101 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124633135","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-06DOI: 10.1109/ICSET53708.2021.9612433
Muhamad Fariz Absl Kamarulazman, Adolfientje Kasenda Olesen, Shahrul Reza Natnan, SAIFUL AMAN Bin HJ SULAIMAN
North Sumatera earthquake on 11 April 2012 has occurred and some of the close countries felt the tremor. The impact was recorded by the Global Navigation Satellite System (GNSS) at the Continuously Operating Reference Station (CORS) throughout Malaysia. This paper investigates the feasibility of using Malaysia continuously operating reference stations (CORS) for short term deformation monitoring and analysis. The data quality and reliability of the GNSS are critical issues in terms of suitability and geological stability. An investigation into appropriate strategies for GNSS data processing and deformation analysis in relation to the most recent International Terrestrial Reference Frame is conducted (ITRF). Double different processing method were used to get the adjusted coordinate in daily solution. The results of deformation analyses, as well as detailed data-processing strategies, are discussed in detail, and some useful conclusions are drawn. The results demonstrate that the deformation analysis derived from regional CORS network data processing is both feasible and effective in practice. In this research the displacement of the CORS stations are between 4.47 – 0.35 magnitude. As evidenced by this example, continuous tracking data from Malaysia's CORS network (MyRTKnet) is a valuable asset that can be used to develop a technically advanced and cost-effective geoscientific infrastructure for deformation monitoring analysis.
{"title":"Displacement of MyRTKNet Stations after M8.6 Earthquake of North Sumatera 2012 using GNSS Data","authors":"Muhamad Fariz Absl Kamarulazman, Adolfientje Kasenda Olesen, Shahrul Reza Natnan, SAIFUL AMAN Bin HJ SULAIMAN","doi":"10.1109/ICSET53708.2021.9612433","DOIUrl":"https://doi.org/10.1109/ICSET53708.2021.9612433","url":null,"abstract":"North Sumatera earthquake on 11 April 2012 has occurred and some of the close countries felt the tremor. The impact was recorded by the Global Navigation Satellite System (GNSS) at the Continuously Operating Reference Station (CORS) throughout Malaysia. This paper investigates the feasibility of using Malaysia continuously operating reference stations (CORS) for short term deformation monitoring and analysis. The data quality and reliability of the GNSS are critical issues in terms of suitability and geological stability. An investigation into appropriate strategies for GNSS data processing and deformation analysis in relation to the most recent International Terrestrial Reference Frame is conducted (ITRF). Double different processing method were used to get the adjusted coordinate in daily solution. The results of deformation analyses, as well as detailed data-processing strategies, are discussed in detail, and some useful conclusions are drawn. The results demonstrate that the deformation analysis derived from regional CORS network data processing is both feasible and effective in practice. In this research the displacement of the CORS stations are between 4.47 – 0.35 magnitude. As evidenced by this example, continuous tracking data from Malaysia's CORS network (MyRTKnet) is a valuable asset that can be used to develop a technically advanced and cost-effective geoscientific infrastructure for deformation monitoring analysis.","PeriodicalId":433197,"journal":{"name":"2021 IEEE 11th International Conference on System Engineering and Technology (ICSET)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121864312","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-06DOI: 10.1109/ICSET53708.2021.9612567
D. Maharani, C. Machbub, P. Rusmin, L. Yulianti
Moving object tracking is the most important component in many computer vision applications. Currently, the ability of computer vision is almost like human vision. Humans can see and track moving objects by looking at notable features such as color, shape, and function. The computer can track moving objects by calculating the characteristics, such as the Histogram of Oriented Gradient (HOG) and grayscale features. These features were used as input in the tracker algorithm. The correlation filter algorithm is extensively used in object tracking applications because of its accuracy and speed. Kernelized Correlation Filters (KCF) is a method that uses correlation for object tracking. The feature fusion is widely used to make tracking more robust. In this paper, the HOG and grayscale features were implemented in the KCF method. Deep Neural Network (DNN) regression was used as a decision feature fusion. With almost similar principle as Non-Maximum Suppression (NMS), where two candidates are detected from overlapping HOG and grayscale features, the region-of-interest (ROI) will be pruned by replacing one ROI to produce a more accurate object candidate. In this study, three TB dataset videos were used for testing, and two videos were used for training. The DNN Regression architecture uses six hidden layers with 512, 256,64,32,16, and 8 nodes. The training accuracy results were 95.76%, with MSE of 9.94 and a loss of 9.93. This research shows that the system can track objects more precisely and robustly with RMSE of 9.38 while achieving 32 FPS.
{"title":"Feature Fusion with Deep Neural Network in Kernelized Correlation Filters Tracker","authors":"D. Maharani, C. Machbub, P. Rusmin, L. Yulianti","doi":"10.1109/ICSET53708.2021.9612567","DOIUrl":"https://doi.org/10.1109/ICSET53708.2021.9612567","url":null,"abstract":"Moving object tracking is the most important component in many computer vision applications. Currently, the ability of computer vision is almost like human vision. Humans can see and track moving objects by looking at notable features such as color, shape, and function. The computer can track moving objects by calculating the characteristics, such as the Histogram of Oriented Gradient (HOG) and grayscale features. These features were used as input in the tracker algorithm. The correlation filter algorithm is extensively used in object tracking applications because of its accuracy and speed. Kernelized Correlation Filters (KCF) is a method that uses correlation for object tracking. The feature fusion is widely used to make tracking more robust. In this paper, the HOG and grayscale features were implemented in the KCF method. Deep Neural Network (DNN) regression was used as a decision feature fusion. With almost similar principle as Non-Maximum Suppression (NMS), where two candidates are detected from overlapping HOG and grayscale features, the region-of-interest (ROI) will be pruned by replacing one ROI to produce a more accurate object candidate. In this study, three TB dataset videos were used for testing, and two videos were used for training. The DNN Regression architecture uses six hidden layers with 512, 256,64,32,16, and 8 nodes. The training accuracy results were 95.76%, with MSE of 9.94 and a loss of 9.93. This research shows that the system can track objects more precisely and robustly with RMSE of 9.38 while achieving 32 FPS.","PeriodicalId":433197,"journal":{"name":"2021 IEEE 11th International Conference on System Engineering and Technology (ICSET)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131286216","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-06DOI: 10.1109/ICSET53708.2021.9612552
Gim Hoy Soong, Chye Cheah Tan
10K Financial reports are submitted by public listed companies to the Security Exchange Commission (SEC) yearly or quarterly. It allows investors to understand strategic planning and directions of the business organization. Although true facts are required to be presented in the reports, it does not prevent companies from using confusing explanations to beautify the organizations current state. Hence, an automated approach to filter out sentiments from the reports is crucial to assist investors in evaluating financial reports. This research paper explores machine learning approaches to conduct sentiment analysis on 10K financial reports. Two different datasets were intended to be used for training the model but only the financial phrase bank dataset was used to produce the final machine learning models. Four machine learning models including fastText, Naïve Bayes Support Vector Machine (NBSVM), Bidirectional Gated Recurrent Units (BiGRU), and Bidirectional Encoder Representations from Transformers (BERT) are trained based on the financial phrase bank dataset. It is discovered that the BERT model performed with the best accuracy while testing the models while the fastText model provided the fastest loading and training time. Conclusion of this research paper shows that different machine learning models in sentiment analysis possess respective advantages and disadvantages and further research can be done with the combination of textual and numerical data in financial reports.
{"title":"Sentiment Analysis on 10-K Financial Reports using Machine Learning Approaches","authors":"Gim Hoy Soong, Chye Cheah Tan","doi":"10.1109/ICSET53708.2021.9612552","DOIUrl":"https://doi.org/10.1109/ICSET53708.2021.9612552","url":null,"abstract":"10K Financial reports are submitted by public listed companies to the Security Exchange Commission (SEC) yearly or quarterly. It allows investors to understand strategic planning and directions of the business organization. Although true facts are required to be presented in the reports, it does not prevent companies from using confusing explanations to beautify the organizations current state. Hence, an automated approach to filter out sentiments from the reports is crucial to assist investors in evaluating financial reports. This research paper explores machine learning approaches to conduct sentiment analysis on 10K financial reports. Two different datasets were intended to be used for training the model but only the financial phrase bank dataset was used to produce the final machine learning models. Four machine learning models including fastText, Naïve Bayes Support Vector Machine (NBSVM), Bidirectional Gated Recurrent Units (BiGRU), and Bidirectional Encoder Representations from Transformers (BERT) are trained based on the financial phrase bank dataset. It is discovered that the BERT model performed with the best accuracy while testing the models while the fastText model provided the fastest loading and training time. Conclusion of this research paper shows that different machine learning models in sentiment analysis possess respective advantages and disadvantages and further research can be done with the combination of textual and numerical data in financial reports.","PeriodicalId":433197,"journal":{"name":"2021 IEEE 11th International Conference on System Engineering and Technology (ICSET)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131372537","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-06DOI: 10.1109/ICSET53708.2021.9612541
P. Netisopakul, Usanisa Taoto
This research constructs and evaluates text generation models created from three different language models, n-gram, a Continuous Bag of Words (CBOW) and gated recurrent unit (GRU), using two training corpora, Berkeley Restaurant (Berkeley) and Alice's Adventures in Wonderland (Alice), and evaluated using two evaluation metrics; perplexity measure and count of grammar errors. The mean perplexities of all three models are comparable for each corpus, the N-gram model produces slightly lower values of perplexity. As for the number of grammatical errors in the Alice corpus, all three models show a slightly higher number of errors than the original corpus. In the Berkeley corpus, the n-gram model had the lowest number of errors, even lower than the original corpus, but the CBOW model had the highest number of errors and the GRU model had the highest number of errors.
{"title":"Word-level Text Generation from Language Models","authors":"P. Netisopakul, Usanisa Taoto","doi":"10.1109/ICSET53708.2021.9612541","DOIUrl":"https://doi.org/10.1109/ICSET53708.2021.9612541","url":null,"abstract":"This research constructs and evaluates text generation models created from three different language models, n-gram, a Continuous Bag of Words (CBOW) and gated recurrent unit (GRU), using two training corpora, Berkeley Restaurant (Berkeley) and Alice's Adventures in Wonderland (Alice), and evaluated using two evaluation metrics; perplexity measure and count of grammar errors. The mean perplexities of all three models are comparable for each corpus, the N-gram model produces slightly lower values of perplexity. As for the number of grammatical errors in the Alice corpus, all three models show a slightly higher number of errors than the original corpus. In the Berkeley corpus, the n-gram model had the lowest number of errors, even lower than the original corpus, but the CBOW model had the highest number of errors and the GRU model had the highest number of errors.","PeriodicalId":433197,"journal":{"name":"2021 IEEE 11th International Conference on System Engineering and Technology (ICSET)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133117648","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：