检索结果-内蒙古大学图书馆

2023 International Seminar on Application for Technology of Information and Communication, iSemantic 2023

作者： Rahmawati, Lailia Rustad, Supriadi Marjuni, Aris Soeleman, Moch Arief Nurtantio Andono, Pulung Pujiono Universitas Dian Nuswantoro Faculty of Computer Science Semarang Indonesia Universitas Darul Ulum Faculty of Informatics Engineering Jombang Indonesia

ISBN: (纸本)9798350339215

This paper discusses a comparison of several methods used for object detection in bad weather conditions, such as the presence of fog that interferes with the object's view. The methods compared in this paper include Faster R-CNN, Yolov3, and SSD (Single Shot Multi-Box Detector). This work contributes to evaluating the best method to overcome the object detection problem in bad weather, which is measured by mean average precision (mAP). The evaluation is executed using video objects that are converted into frames. Based on experimental results, the Faster R-CNN method is superior in terms of the mAP value where the Faster R-CNN is 87%, using SSD is 48%, and Yolov3 is 31%. Faster R-CNN provides better performance than the other two methods because the basic idea of FRCNN is the use of a Region Proposal Network (RPN) in a feature map which aims to propose many objects that can be identified in a particular image. © 2023 IEEE.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Ottoman Optical Character Recognition with deep neural networks

Journal of the Faculty of Engineering and Architecture of Ga...

引用

Journal of the faculty of engineering and Architecture of Gazi University 2023年第4期38卷 2579-2593页

作者： Dölek, İshak Kurt, Atakan Mina R&D Informatics Entertech Office:322 Istanbul- University Cerrahpaşa Istanbul34320 Turkey Department of Computer Engineering Faculty of Engineering Istanbul University Cerrahpaşa Istanbul34320 Turkey

In this paper, we present a web-based optical character recognition (OCR) system that converts images of Ottoman documents printed with naskh font into text using CNN+RNN-based deep neural network models. For training, three datasets - original, synthetic, and hybrid - were prepared and three different OCR models were created. The original data set consists of 1,000 pages and the synthetic data set consists of 23,000 pages. Hybrid data set contains both. The trained models were compared with Tesseract's Arabic and Persian, Google Docs' Arabic, Abby FineReader's Arabic, and Miletos OCR model/tools with a 21-page test set. The comparison was made with 3 different texts (raw, normalized, and joined) and using 3 different criteria (character, ligature, and word recognition). The *** Hybrid model produced significantly better results than the others with 88.86% raw, 96.12% normalized, and 97.37% joined accuracy in character recognition;80.48% raw, 91.60% normalized, and 97.37% joined accuracy in ligature recognition;and 44.08% raw and 66.45% normalized accuracy in word recognition. To investigate the effects of the characteristics of the alphabet on OCR, character, ligature, and word frequency analyses of Ottoman was performed. In this analysis, the characters in the alphabet were grouped according to distinctive features such as connectedness, letter body, position and number of dots, type of character, and source language;and frequencies and recognition accuracies were examined for each group. OCR results are also reported for each character. © 2023 Gazi Universitesi Muhendislik-Mimarlik. All rights reserved.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Factors Influencing Knowledge Sharing in Prison: A Review and AHP Analysis 6

Factors Influencing Knowledge Sharing in Prison: A Review an...

引用

6th International Conference on computer and informatics engineering, IC2IE 2023

作者： Imandeka, Ejo Hidayanto, Achmad Nizar Suhartanto, Heru Pidanic, Jan University of Indonesia Faculty of Computer Science Depok Indonesia University of Pardubice Faculty of Electrical Engineering and Informatics Pardubice Czech Republic

ISBN: (纸本)9798350345162

Knowledge sharing has been recognized as the most critical factor in successful knowledge management. Knowledge sharing in prison is a complicated topic that frequently faces resource restrictions, stigma and isolation, communication hurdles, and other security concerns. Moreover, resources for knowledge sharing in a secure environment are still few. This research investigates the critical factors for knowledge sharing in prisons. This research uses the Kitchenham approach with five reputable publishers from 2018 to 2023. There are 1284 articles in the initial phase and 25 in the final phase. Critical factors were ranked using Analytical Hierarchy Process (AHP) with the Triple-helix approach to the representatives of academics, policymakers, and the community in the Indonesian context. The critical knowledge-sharing factors are leadership support, organizational culture, organizational structure, incentives, and individual beliefs. However, acknowledging the critical factors can aid in improving the knowledge-sharing system in prison and changing the criminal justice system. © 2023 IEEE.

关键词： analytic hierarchy process critical factors knowledge sharing prison

来源：评论

学校读者我要写书评

暂无评论

Facial Palsy Detection Through Changes in Facial Muscle Functionality Using CNN Algorithm 6

Facial Palsy Detection Through Changes in Facial Muscle Func...

引用

6th International Conference on computer and informatics engineering, IC2IE 2023

作者： Sari, Chindy Aulia Bachtiar, Fitra Abdurrachman Muflikhah, Lailil Widayati, Aris University of Brawijaya Faculty of Computer Science Informatics Engineering Department Malang Indonesia University of Brawijaya Faculty of Medicine Malang Indonesia

ISBN: (纸本)9798350345162

Facial palsy is a common condition in the medical field that results from facial nerve disorders, leading to weakened facial motor functions. Detecting facial palsy typically involves observing the functional ability of facial muscles, a process currently performed manually by healthcare professionals relying on visual cues. However, manual detection is prone to errors, this needs for early and automated facial palsy detection. Overall, our research aims to improve the detection of facial palsy by leveraging deep learning techniques. By automating the detection process, we hope to provide a more accurate and efficient approach for early identification of this condition. In this research, we aim to develop a facial palsy detection system using the Convolutional Neural Network (CNN) algorithm. We utilize three datasets, namely CK+, JAFFE, and YouTube Facial Palsy, which are organized into normal and dysfunctional classes. The dataset is divided into training and testing data in an 80:20 ratio, and preprocessing techniques such as image resizing and grayscale conversion are applied. To find the optimal parameters for our model, we employ k-fold cross validation during the training process. The evaluation results show that the best model achieved optimal performance with the following parameters: 3 hidden layers, a fully connected layer (fc)-1 with a hidden size of 128, fc-2 with a hidden size of 16, and fc-3 with a hidden size of 32. The optimizer used was Adam, with a learning rate of 0.001, a batch size of 64, and trained for 10 epochs. This best model attained an average training accuracy of 0.958, testing accuracy of 1, recall of 1, and precision of 0.99. © 2023 IEEE.

关键词： CNN deep learning face detection facial palsy

来源：评论

学校读者我要写书评

暂无评论

Combination of CRITIC and PROMETHEE II Methods as An Early Detection of Potential Dropout Students 6

Combination of CRITIC and PROMETHEE II Methods as An Early D...

引用

6th International Conference on Information and Communications Technology, ICOIACT 2023

作者： Yusa, Mochammad Coastera, Funny Farady Rahmadani, Deni Sari, Julia Purnama Kusrini, Kusrini Hartanto, Anggit Dwi University of Bengkulu Faculty of Engineering Department of Informatics Bengkulu Indonesia University of Bengkulu Faculty of Engineering Department of Information System Bengkulu Indonesia Universitas Amikom Yogyakarta Magister of Informatics Engineering Faculty of Computer Science Yogyakarta Indonesia Universitas Amikom Yogyakarta Faculty of Computer Science Department of Information System Yogyakarta Indonesia

ISBN: (纸本)9798350315639

Dropout (DO) has negative implications for individuals and educational institutions. The field of education data mining (EDM) offers valuable contributions to prevent dropout cases and improve student retention. This study focuses on implementing EDM to identify students at risk of dropping out by integrating academic activity criteria with real-time data sources. Previous studies have encountered limitations in classification and clustering techniques, with classification relying on labeled training data (supervised learning) and clustering constrained by attribute or criteria weight (unsupervised learning). This study aims to address these limitations by combining the PROMETHEE II ranking approach and the CRITIC weighting method, producing a list of students with dropout potential. The results can aid stakeholders in implementing preventive measures. The study involves several steps: 1) Collecting data through integration with academic portals;2) Filtering and selecting criteria for student academic activities;3) Transforming and normalizing the data;4) Applying automatic weighting using CRITIC;5) Calculating criteria rankings using the PROMETHEE II technique;and 6) Validating the ranking outcomes. © 2023 IEEE.

关键词： Decision support systems

来源：评论

学校读者我要写书评

暂无评论

Scheduling Tool for Deterministic Communication in Distributed Real-Time Systems 10

Scheduling Tool for Deterministic Communication in Distribut...

引用

10th Workshop on Software Quality Analysis, Monitoring, Improvement, and Applications, SQAMIA 2023

作者： Mach, Jan Tašková, Valéria Kohútka, Lukáš Institute of Computer Engineering and Applied Informatics Faculty of Informatics and Information Technologies Slovak University of Technology Bratislava Slovakia Institute of Informatics Information Systems and Software Engineering Faculty of Informatics and Information Technologies Slovak University of Technology Bratislava Slovakia

A network with critical data streams, where the timing of incoming and outgoing data is a necessity, is called a deterministic network. These networks are mostly used in association with real-time systems that use periodic hard-real-time data which need to be scheduled. The algorithm proposed in this paper is focused on achieving the lowest possible latency within the specified network topology depending on the physical parameters of the network and the periodic requirements of the communicating nodes. The algorithm allows to add priorities to the data flow between nodes and also the option to specify if preemption is allowed within the network, since not all devices support it. The result is a schedule of hard-real-time data communication within a specified network topology where all requirements of communicating nodes are met if such a solution is possible. © 2023 Copyright for this paper by its authors.

关键词： deadline deterministic scheduling Ethernet scheduling algorithm time-sensitive networking

来源：评论

学校读者我要写书评

暂无评论

A distributed B+Tree indexing method for processing range queries over streaming data

引用

Cluster Computing 2024年第2期27卷 1251-1274页

作者： Safaee, Shahab Mirabi, Meghdad Rahmani, Amir Masoud Safaei, Ali Asghar Department of Computer Engineering Faculty of Engineering South Tehran Branch Islamic Azad University Tehran Iran Future Technology Research Center National Yunlin University of Science and Technology Yunlin Douliou Taiwan Department of Medical Informatics Faculty of Medical Sciences Tarbiat Modares University Tehran Iran

A data stream exhibits as a massive unbounded sequence of data elements continuously generated at a high rate. Stream databases raise new challenges for query processing due to both the streaming nature of data which constantly changes over time and the wider range of queries submitted by the user when compared with the traditional databases. In this paper, we propose a system architecture which includes components for both distributed indexing of streaming data and distributed processing of range queries on streaming data. Instead of creating a large and centralized B+Tree index structure, we create a set of small B+Tree indexes in such a way that a B+Tree index can be created for every partition of streaming data. We also design a distributed range search algorithm which can be used by each individual machine inside a Spark cluster to independently process range queries on each partition of streaming data. By exploiting the proposed system architecture, the process of indexing of streaming data and the process of querying over streaming data can be performed in a distributed and parallel manner. By performing several experiments, we demonstrate that our proposed indexing method is scalable and efficient for processing range queries on streaming data compared to the existing centralized B+Tree indexing methods and therefore, it can be used for applications involving data streams with a large volume of data elements and a large number of range queries. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2023.

关键词： Query processing

来源：评论

学校读者我要写书评

暂无评论

Bad channel detection in EEG recordings

引用

Current Directions in Biomedical engineering 2022年第2期8卷 257-260页

作者： Komosar, Milana Fiedler, Patrique Haueisen, Jens Institute of Biomedical Engineering and Informatics Faculty of Computer Sciences and Automation Technische Universität Ilmenau Ilmenau Germany

Electroencephalography (EEG) is widely used in clinical applications and basic research. Dry EEG opened the application area to new fields like self-application during gaming and neurofeedback. While recording, the signals are always affected by artefacts. Manual detection of bad channels is the gold standard in both gel-based and dry EEG but is time-consuming. We propose a simple and robust method for automatic bad channel detection in EEG. Our method is based on the iterative calculation of standard deviations for each channel. Statistical measures of these standard deviations serve as indications for bad channel detection. We compare the new method to the results obtained from the manually identified bad channels for EEG recordings. We analysed EEG signals during resting state with eyes closed and datasets with head movement. The results showed an accuracy of 99.69 % for both gel-based and dry EEG for resting state EEG. The accuracy of our new method is 99.38 % for datasets with the head movement for both setups. There was no significant difference between the manual gold standard of bad channel identification and our iterative standard deviation method. Therefore, the proposed iterative standard deviation method can be used for bad channel detection in resting state and movement EEG recordings. © 2022 The Author(s), published by De Gruyter.

关键词： Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

EEG emotion recognition framework based on invariant wavelet scattering convolution network

引用

Journal of Ambient Intelligence and Humanized Computing 2024年第4期15卷 2181-2199页

作者： Elrefaiy, Ahmed Tawfik, Nahed Zayed, Nourhan Elhenawy, Ibrahim Department of Computer Science Faculty of Computers and Informatics Zagazig University Zagazig44519 Egypt Computers and Systems Department Electronics Research Institute Cairo Egypt Mechanical Engineering Department The British University in Egypt Cairo Egypt

EEG signals for real-time emotion identification are crucial for affective computing and human-computer interaction. The current emotion recognition models, which rely on a small number of emotion classes and stimuli like music and images in controlled lab conditions, have poor ecological validity. Furthermore, identifying relevant EEG signal features is crucial for efficient emotion identification. According to the complexity, non-stationarity, and variation nature of EEG signals, which make it challenging to identify relevant features to categorize and identify emotions, a novel approach for feature extraction and classification concerning EEG signals is suggested based on invariant wavelet scattering transform (WST) and support vector machine algorithm (SVM). The WST is a new time-frequency domain equivalent to a deep convolutional network. It produces scattering feature matrix representations that are stable against time-warping deformations, noise-resistant, and time-shift invariant existing in EEG signals. So, small, difficult-to-measure variations in the amplitude and duration of EEG signals can be captured. As a result, it addresses the limitations of the previous feature extraction approaches, which are unstable and sensitive to time-shift variations. In this paper, the zero, first, and second order features from DEAP datasets are obtained by performing the WST with two deep layers. Then, the PCA method is used for dimensionality reduction. Finally, the extracted features are fed as inputs for different classifiers. In the classification step, the SVM classifier is utilized with different classification algorithms such as k-nearest neighbours (KNN), random forest (RF), and AdaBoost classifier. This research employs a principal component analysis (PCA) approach to reduce the high dimensionality of scattering characteristics and increase the computational efficiency of our classifiers. The proposed method is performed across four different emotional classific

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Novel replaceable EEG electrode system

引用

Current Directions in Biomedical engineering 2022年第2期8卷 249-252页

作者： Warsito, Indhika Fauzhan Fiedler, Patrique Komosar, Milana Haueisen, Jens Institute of Biomedical Engineering and Informatics Faculty of Computer Science and Automation Technische Universität Ilmenau Ilmenau Germany

Due to the direct contact between electrode and scalp, dry EEG electrodes are exposed to increased mechanical wear compared to conventional gel-based electrodes. However, state-of-the-art commercial cap systems commonly use permanently fixated electrodes which can lead to downtime of the EEG cap during professional repair and replacement as well as reduced overall lifetime. An easily replaceable EEG electrode would furthermore improve hygiene, especially for newborn and infant applications. We propose a novel replaceable electrode system, consisting of an electrode holder, a snap top, a contact ring fixated inside the electrode holder, and a replaceable electrode. The production process consists of 3D printing, silicone molding, resin casting, and electroless plating. The replaceable electrode system is integrated into a multichannel EEG cap system. A verification study is conducted with 30 volunteers. The operators experienced that the new electrode holder eases adjustment of the electrode to have proper contact with the scalp. During the study, defective electrodes can be replaced without a soldering process. Furthermore, all electrodes stayed in the holder and did not fall off the cap for the whole session. In conclusion, the novel replaceable electrode system is suitable for EEG measurements. © 2022 The Author(s), published by De Gruyter.

关键词： Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：