检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Guo, Xiaotong Yang, Deqian Wang, Dan Zhao, Haochen Li, Yuan Sui, Zhilin Zhou, Tao Zhang, Lijun Meng, Yanda State Key Laboratory of Computer Science Institute of Software Chinese Academy of Sciences China Department of Thoracic Surgery National Clinical Research Center for Cancer Cancer Hospital Shenzhen Hospital China School of Intelligent Science and Technology Hangzhou Institute for Advanced Study University of Chinese Academy of Sciences Hangzhou310024 China R&D Center Guangxi Huayi Artificial Intelligence Medical Technology Co. Ltd. China School of Computer Science and Engineering Beihang University Beijing China Department of Computer Science University of Exeter Exeter United Kingdom Department of Cardiovascular & Metabolic Medicine University of Liverpool Liverpool United Kingdom Guangzhou Jiayi Software Technology Co. Ltd. China

Accurate segmentation of pulmonary structures is crucial in clinical diagnosis, disease study, and treatment planning. Significant progress has been made in deep learning-based segmentation techniques, but most require much labeled data for training. Consequently, developing precise segmentation methods that demand fewer labeled datasets is paramount in medical image analysis. The emergence of pre-trained vision-language foundation models, such as CLIP, recently opened the door for universal computer vision tasks. Exploiting the generalization ability of these pre-trained foundation models on downstream tasks, such as segmentation, leads to unexpected performance with a relatively small amount of labeled data. However, exploring these models for pulmonary artery-vein segmentation is still limited. This paper proposes a novel framework called Language-guided self-adaptive Cross-Attention Fusion Framework. Our method adopts pre-trained CLIP as a strong feature extractor for generating the segmentation of 3D CT scans, while adaptively aggregating the cross-modality of text and image representations. We propose a s pecially designed adapter module to fine-tune pre-trained CLIP with a self-adaptive learning strategy to effectively fuse the two modalities of embeddings. We extensively validate our method on a local dataset, which is the largest pulmonary artery-vein CT dataset to date and consists of 718 labeled data in total. The experiments show that our method outperformed other state-of-the-art methods by a large margin. Our data and code will be made publicly available upon acceptance. Copyright © 2025, The Authors. All rights reserved.

关键词： Visual languages

来源：评论

学校读者我要写书评

暂无评论

User Acceptance of New HCI Technologies

User Acceptance of New HCI Technologies

引用

International IEEE Conference and Workshop: Óbuda on Electrical and Power engineering (CANDO-EPE)

作者： Jozsef Katona Attila Kovari Kandó Kálmán Faculty of Electrical Engineering Institute of Electronics and Communication Systems Obuda University Budapest Hungary Department of Software Development and Application Institute of Computer Engineering University of Dunaujvaros Dunaujvaros Hungary GAMF Faculty of Engineering and Computer Science John Von Neumann University Kecskemét Hungary Faculty of Computer Science Institute of Digital Technology Eszterházy Károly Catholic University Eger Hungary Department of Natural Sciences Institute of Engineering University of Dunaujvaros Dunaujvaros Budapest Hungary

Human-computer interaction (HCI) is an evolving field of research that focuses on understanding and improving the communication and interaction between humans and computers. Over the past decades, we have seen many significant advances in this field, which have contributed to the widespread adoption and integration of technology into our daily lives. HCI research and development aims to design information technology systems to meet the needs and preferences of users. Usability, efficiency, accessibility and user satisfaction are important considerations in HCI design. This paper presents the results of a pilot study on user acceptance of HCIs. The results show that users are positive about and willing to use HCIs.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Neural Network Pruning by Cooperative Coevolution

arXiv

引用

arXiv 2022年

作者： Shang, Haopu Wu, Jia-Liang Hong, Wenjing Qian, Chao State Key Laboratory for Novel Software Technology Nanjing University Nanjing210023 China Department of Computer Science and Engineering Southern University of Science and Technology Shenzhen518055 China

Neural network pruning is a popular model compression method which can significantly reduce the computing cost with negligible loss of accuracy. Recently, filters are often pruned directly by designing proper criteria or using auxiliary modules to measure their importance, which, however, requires expertise and trial-and-error. Due to the advantage of automation, pruning by evolutionary algorithms (EAs) has attracted much attention, but the performance is limited for deep neural networks as the search space can be quite large. In this paper, we propose a new filter pruning algorithm CCEP by cooperative coevolution, which prunes the filters in each layer by EAs separately. That is, CCEP reduces the pruning space by a divide-and-conquer strategy. The experiments show that CCEP can achieve a competitive performance with the state-of-the-art pruning methods, e.g., prune ResNet56 for 63.42% FLOPs on CIFAR10 with −0.24% accuracy drop, and ResNet50 for 44.56% FLOPs on ImageNet with 0.07% accuracy drop. Copyright © 2022, The Authors. 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

Design and Implementation an Open Source Web Based Clearance System for Iraqi Educational Institutions 2

Design and Implementation an Open Source Web Based Clearance...

引用

2nd Annual International Conference on Information and sciences, AiCIS 2020

作者： Ramadhan, Rebaz N. Manguri, Kamaran H. Manguri, Balen H. Koya University Software Engineering Department Koya Iraq University of Raparin Computer Science Department Rania Iraq Soran University Computer Science Department Soran Iraq

ISBN: (纸本)9781728191690

Clearance is a mandatory process in almost every university. Either graduated or undergraduate students must undertake the procedure to grant the permission to participate in the final exams. In addition to this, students need stamped clearance paper to receive their final exam results. Over the last decades, clearance has been carried out manually. With the manual clearance system, students endure a long and exhausting process of signatures and stamps. In the Iraqi educational system, the process involves visiting many departments including library, registrar, accounting, accommodation, laboratory, storage, sport and artistic activity departments, and the dean's office;only to have the clearance certified. The aim of this project is to propose an open source clearance system. In simple words, the proposed system will have several advantages over the old system for instance time and effort, cost, speed, and data safety. Languages such as HTML, CSS, and JavaScript are used to design and implement the front-end side. On the other hand, its back-end is created by the help of PHP scripts and MySQL for the database. Generally, the system has four equally important major tasks to be fulfilled in order to have the clearance. The tasks are registration, unit administrator (to confirm a clearance status as Cleared/Not Cleared), Exam committee (to check whether student's clearance status), and finally, a system administrator (to manage the entire system). Furthermore, system admin can add or remove any unit administrator or any student to/from the system. This proposed system is available online and can be found at the following link: ***/hvwNY © 2020 IEEE.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

Diagnosis and Identification of Citrus Canker Growth Rate Using Machine Learning

Diagnosis and Identification of Citrus Canker Growth Rate Us...

引用

Pattern Recognition and Machine Learning (PRML), IEEE International Conference on

作者： Naureen Zainab Hammad Afzal Fahim Arif Abdul Ghafoor Naima Iltaf Muhammad Zakria Muhammad Javed Hyder Rafia Ahsan Department of Computer Software Engineering National University of Sciences and Technology Islamabad Pakistan National Agricultural Research Centre (NARC) Park Road Crop Diseases Research Institute (CDRI) Islamabad Pakistan Department of Mechanical Engineering Capital University of Science and Technology (CUST) Islamabad Pakistan

Citrus canker is a bacterial disease caused by Xanthomonas citri pv citri. Mainly the research in the field of Citrus canker has been done on fruits; early detection of disease on leaves can help in adopting preventive measures at early stage. The present study focuses on the detection of Citrus canker in leaves and classification of six different growth stages/levels of disease: water soaking, yellow initiation, chlorosis, and blister formation, canker infection (50%), canker infection (100%). The primary dataset was developed by taking images of citrus leaves infected by X. citri for the specified time. The proposed strategy is based on the detection of the lesion spot on citrus leaves and the classification of citrus disease stages/level was done to evaluate the growth rate of citrus canker through six distinct growth levels and evaluate affected area of leaf with each passing day. The entire processing includes Pre-processing, Segmentation, Features extraction and perform Classification using four different classifiers which include SVM, KNN, Naive Bayes, and Neural Network. The efficiency of different classifiers was 96.77% by SVM, followed by KNN 90%, NB 84% and NN 82% respectively. The lowest execution time was achieved by NB 0.011s, followed by KNN 0.015s, SVM 0.021s, and N.N. 0.216s.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Refining the Unseen: Self-supervised Two-stream Feature Extraction for Image Quality Assessment

Refining the Unseen: Self-supervised Two-stream Feature Extr...

引用

IEEE International Conference on Data Mining (ICDM)

作者： Yiwei Lou Yanyuan Chen Dexuan Xu Doudou Zhou Yongzhi Cao Hanpin Wang Yu Huang Key Laboratory of High Confidence Software Technologies (Peking University) Ministry of Education School of Computer Science Peking University Beijing China School of Software & Microelectronics Peking University Beijing China Department of Biostatistics Harvard T.H. Chan School of Public Health Boston USA National Engineering Research Center for Software Engineering Peking University Beijing China

The inadequacy of labeled datasets for image quality assessment has led to the development and popularity of self-supervised approaches. However, most existing self-supervised methods primarily focus on content and fidelity features extracted with convolutional neural networks, overlooking the crucial importance of structural features in quality assessment. To address this problem, we present a novel self-supervised two-stream feature extraction and representation approach. In our approach, the first stream leverages a contrastive learning framework to extract image fidelity features, while the second stream emphasizes structural features by incorporating an attention mechanism. This innovative combination results in a comprehensive feature representation for quality assessment. Moreover, our proposed method facilitates transfer learning, allowing the pre-trained two-stream model in the source domain to be seamlessly applied to target domains for quality regression. This compatibility with transfer learning enhances the adaptability and generalization of the model. Extensive experiments are carried out on three synthetic distortion datasets to validate the effectiveness of our approach. The results demonstrate that our work not only competes with state-of-the-art self-supervised methods but also outperforms some supervised approaches.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Robust Image Ordinal Regression with Controllable Image Generation

arXiv

引用

arXiv 2023年

作者： Cheng, Yi Ying, Haochao Hu, Renjun Wang, Jinhong Zheng, WenHao Zhang, Xiao Chen, Danny Wu, Jian School of Software Technology Zhejiang Univerisity China School of Public Health Zhejiang University China Alibaba Group China College of Computer Science and Technology Zhejiang University China School of Computer Science and Technology Shandong University China Department of Computer Science and Engineering University of Notre Dame United States Second Affiliated Hospital School of Medicine Zhejiang University China

Image ordinal regression has been mainly studied along the line of exploiting the order of categories. However, the issues of class imbalance and category overlap that are very common in ordinal regression were largely overlooked. As a result, the performance on minority categories is often unsatisfactory. In this paper, we propose a novel framework called CIG based on controllable image generation to directly tackle these two issues. Our main idea is to generate extra training samples with specific labels near category boundaries, and the sample generation is biased toward the less-represented categories. To achieve controllable image generation, we seek to separate structural and categorical information of images based on structural similarity, categorical similarity, and reconstruction constraints. We evaluate the effectiveness of our new CIG approach in three different image ordinal regression scenarios. The results demonstrate that CIG can be flexibly integrated with off-the-shelf image encoders or ordinal regression models to achieve improvement, and further, the improvement is more significant for minority categories. Copyright © 2023, The Authors. All rights reserved.

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

Personalization of Job Circular Announcement to Applicants Using Decision Tree Classification Algorithm

Personalization of Job Circular Announcement to Applicants U...

引用

2020 IEEE International Conference for Innovation in Technology, INOCON 2020

作者： Javed Mehedi Shamrat, F.M. Tasnim, Zarrin Ghosh, Pronab Majumder, Anup Hasan, Md. Zahid Daffodil International University Department of Software Engineering Dhaka Bangladesh Jahangirnagar University Department of Computer Science and Engineering Dhaka Bangladesh

ISBN: (纸本)9781728197449

While looking for jobs, job seekers like to register themselves in different job portal that has a huge number of job circulars. As from all the job circulars, it is easier to look for the job of their liking. At the same time, these job portals gather some information on the jobseekers that they use to apply to different jobs. Whenever a new job circular is posted in the portal, it is advertised to the job seekers through email notification. To make the email notification more personalized and relevant to the job seekers, a decision tree algorithm is implemented. This decision tree makes the decision of which job circular should be advertised to which job seekers based on the job seeker's field of interest, expected salary and experience. This helps jobseekers relate to the notification and be more engaged in the job portal. © 2020 IEEE.

关键词： Decision trees

来源：评论

学校读者我要写书评

暂无评论

Application of soft computing algorithms for hybrid modular multilevel inverters

引用

Measurement: Sensors 2024年 31卷

作者： Gireesh Kumar, Devineni Venkata Sireesha, Nagineni Bhoopal, Neerudi Kumar Gatla, Ranjith Kotb, Hossam AboRas, Kareem M. Elrashidi, Ali Alqarni, Mohammed Yasin Ghadi, Yazeed Oubelaid, Adel Department of Electrical & Electronics Engineering B V Raju Institute of Technology Narsapur502313 India Department of Information Technology Institute of Aeronautical Engineering Hyderabad500043 India Department of Electrical & Electronics Engineering Institute of Aeronautical Engineering Hyderabad500043 India Department of Electrical Power and Machines Faculty of Engineering Alexandria University Alexandria21544 Egypt Electrical Engineering Department University of Business and Technology Ar Rawdah Jeddah23435 Saudi Arabia Engineering Mathematics Department Faculty of Engineering Alexandria University Alexandria21544 Egypt Department of Computer Science and Software Engineering Al Ain University Abu Dhabi15322 United Arab Emirates Laboratoire de Technologie Industrielle et de L'Information Faculté de Technologie Université de Bejaia Bejaia06000 Algeria

This study utilized Harris Hawk optimization to analyze the harmonic distortions of a symmetric capacitor based multilevel converter. The proposed multilevel converter uses identical and symmetric DC sources in its input. The proposed topology for a multilevel converter uses fewer switches than typical cascaded designs. It can output 9-level outputs and can be scaled up to higher levels. The proposed multilevel converter can be improved by implementing a low frequency pulse width modulation technique to reduce the stress on the switches. The optimization of the switching angles using Harris Hawk was able to solve the non-linear aspects of the Selective Harmonic Elimination Pulse Width Modulation. The output voltage's THD was compared with that of other optimization methods, such as ant colony and particle swarm. The comparison shows that the output THD of the Harris Hawk optimizer is lower than that of other methods, which is about 5 %, according to the standards of IEEE-519. © 2023

关键词： Pulse width modulation

来源：评论

学校读者我要写书评

暂无评论

The Efficiency in Controlling and Monitoring a Poultry Farm based on Internet of Things (IoT)

The Efficiency in Controlling and Monitoring a Poultry Farm ...

引用

IEEE International Conference on Automatic Control and Intelligent Systems (I2CACIS)

作者： Md Gapar Md Johar Fatria Jumara Adha Asif Iqbal Hajamydeen Lilysuriazna Raya Mohammed Hazim Alkawaz Software Engineering and Digital Innovation Centre Management and Science University Shah Alam Selangor Malaysia School of Graduates Studies Management and Science University Shah Alam Selangor Malaysia Artificial Intelligence and Cyber Security Centre Management and Science University Shah Alam Selangor Malaysia Faculty of Information Sciences and Engineering Management and Science University Shah Alam Selangor Malaysia Department of Computer Science College of Education for Pure Science University of Mosul Mosul Nineveh Iraq

ISBN: (数字)9798350372106

ISBN: (纸本)9798350372113

Every year technology continues to develop, especially in the field of animal husbandry. As demonstrated by the 3.6% increase in chicken consumption in developing nations, it is predicted that both poultry production and consumption will rise between 2005 until 2030. As a result, the chicken business needs to be able to fulfill customer demand. Poultry, particularly chickens, are sensitive to changes in temperature, humidity, and ammonia gas inside the coop, all of which can make the chickens sick. As a result, research is being done to create a server system and develop a framework utilizing IoT technologies. By utilizing Internet of Things (IoT) technology, efficiency in monitoring and controlling chickens can increase production. There are many farmers who still use traditional methods, especially small and medium-scale farmers. This system increases the level of measurement accuracy in the chicken coop using five modules and the device used is flexible so that farmers can move it and place it anywhere. using this system, farmers can manage chicken farming more effectively and measurably, which farmers can manage simply by monitoring via a website that can be accessed anytime and anywhere.

关键词： Temperature sensors Microcontrollers System performance Production Real-time systems Manufacturing Internet of Things

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：