检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Wu, Xu Hou, XianXu Lai, Zhihui Zhou, Jie Zhang, Ya-Nan Pedrycz, Witold Shen, Linlin The Computer Vision Institute College of Computer Science and Software Engineering Shenzhen University Shenzhen518060 China Shenzhen Institute of Artificial Intelligence and Robotics for Society Shenzhen518060 China Guangdong Key Laboratory of Intelligent Information Processing Shenzhen University Shenzhen518060 China School of AI and Advanced Computing Xi’an Jiaotong-Liverpool University China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University SZU Branch Shenzhen Institute of Artificial Intelligence and Robotics for Society Guangdong Shenzhen518060 China The Department of Electrical & Computer Engineering University of Alberta University of Alberta Canada

Low-light image enhancement (LLIE) aims to improve low-illumination images. However, existing methods face two challenges: (1) uncertainty in restoration from diverse brightness degradations;(2) loss of texture and color information caused by noise suppression and light enhancement. In this paper, we propose a novel enhancement approach, CodeEnhance, by leveraging quantized priors and image refinement to address these challenges. In particular, we reframe LLIE as learning an image-to-code mapping from low-light images to discrete codebook, which has been learned from high-quality images. To enhance this process, a Semantic Embedding Module (SEM) is introduced to integrate semantic information with low-level features, and a Codebook Shift (CS) mechanism, designed to adapt the pre-learned codebook to better suit the distinct characteristics of our low-light dataset. Additionally, we present an Interactive Feature Transformation (IFT) module to refine texture and color information during image reconstruction, allowing for interactive enhancement based on user preferences. Extensive experiments on both real-world and synthetic benchmarks demonstrate that the incorporation of prior knowledge and controllable information transfer significantly enhances LLIE performance in terms of quality and fidelity. The proposed CodeEnhance exhibits superior robustness to various degradations, including uneven illumination, noise, and color distortion. Copyright © 2024, 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

A Heuristic Sparsification Method to Improve the Performance of Neural Networks on Autism Spectrum Disorder Identification Tasks

A Heuristic Sparsification Method to Improve the Performance...

引用

International Conference on Information science and Technology (ICIST)

作者： Fangyu Zhang Yi Pan Yanjie Wei Jin Liu Zhenli Lin Centre for High Performance Computing Chinese Academy of Sciences Shenzhen Institute of Advanced Technology Shenzhen China Shenzhen Key Laboratory of Intelligent Bioinformatics Chinese Academy of Sciences Shenzhen Institute of Advanced Technology Shenzhen China School of Computer Science and Engineering Central South University Changsha China Department of Ophthalmology Shenzhen University General Hospital Shenzhen China

Deep learning (DL) has been widely used in the detection of neurological diseases such as Autism spectrum disorder (ASD). Most studies in this field manually designed and adjusted the hyperparameters of DL models, but due to the huge search space, few studies have attempted to optimize the topology of neural networks which is another important factor related to the performance of DL models. This paper proposes a heuristic sparsification method named the Evolutionary Connection Sparsification Algorithm (ECSA) which is used to optimize the topology of fully connected neural networks (FC-NNs) and generate the corresponding sparsely connected neural networks (SC-NNs). The algorithm is based on the workflow of the genetic algorithm but with improved coding strategy, fitness evaluation approach, crossover and mutation operator, and including an effective convergence acceleration strategy. Experiments show that the SC-NNs generated by ECSA significantly outperform the original FC-NNs in terms of accuracy, F1-score, and AUC, and the model complexities are greatly reduced.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Artificial Intelligence and Machine Learning: Developing the Future of Text Analytics and Natural Language Processing (NLP)

Artificial Intelligence and Machine Learning: Developing the...

引用

Smart Technologies for Sustainable Development Goals (ICSTSDG), International Conference on

作者： N.K.Senthil Kumar Rajasekaran. A K. Saraswathi Srinadh Unnava V.Samuthira Pandi Lakshmi Priya J Department of Computer Science & Engineering School of Computing Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology Chennai Tamil Nadu India Department of Computer Science and Engineering Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences (SIMATS) Chennai Tamil Nadu India Department of Electronics and Instrumentation Engineering SCSVMV Deemed to be University Kanchipuram Tamil Nadu India Department of Information Technology Sasi Institute of Technology & Engineering West Godavari(DT) Andhra Pradesh India Centre for Advanced Wireless Integrated Technology Chennai Institute of Technology Chennai Tamil Nadu India Department of ECE Saveetha Engineering College Chennai Tamil Nadu India

ISBN: (数字)9798331516284

ISBN: (纸本)9798331516291

Natural language processing (NLP) and text analytics are two areas within the fast developing field of machine learning (ML) and artificial intelligence (AI). A new age of machine language understanding is dawning, and this research delves deeply into how AI and ML are improving NLP's capabilities and setting the stage for it. From traditional rule-based natural language processing systems to the rise of sophisticated ML algorithms capable of learning and adapting from massive datasets, we follow their evolution. The groundbreaking work done by deep learning models—including RNNs, LSTMs, and the groundbreaking transformer architecture—is highlighted in particular for the remarkable improvements in accuracy and efficiency of language tasks. Moreover, the article delves into how pre-trained language models like BERT and the incorporation of transfer learning have revolutionised the discipline, enabling the lightning-fast creation of NLP applications in a wide range of domains. In addition, we discuss the difficulties that come with these developments, such as the importance of privacy in data-driven ML methods, the ethical considerations around AI's bias and fairness, and the necessity for stronger and more broadly applicable models. A prospective view on how AI and ML may transform text analytics and natural language processing is included in the study's conclusion. More open and reliable natural language processing (NLP) systems may be possible in the future because to initiatives like explainable AI and multimodal learning. With the further development of AI and ML comes the possibility of revolutionising many industries and reshaping human-computer interaction via the processing of natural language at a high level of sophistication. The paper highlights these enormous opportunities.

关键词： Human computer interaction Deep learning Ethics Data privacy Adaptation models Explainable AI Transfer learning Transforms Natural language processing Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

Collaborative Defense with Multiple USVs and UAVs Based on Swarm Intelligence

引用

Journal of Shanghai Jiaotong university(science) 2020年第1期25卷 51-56页

作者： WU Xing LIU Yuan XIE Shaorong GUO Yike School of Computer Engineering and Science Shanghai UniversityShanghai 200444China Shanghai Institute for Advanced Communication and Data Science Shanghai 200444China Department of Computing Imperial College LondonLondon SW72AZUnited Kingdom

Modern defense systems are developing towards *** and automation,which include the collaborative defense system on the sea between multiple unmanned surface vehicles(USVs)and unmanned aerial vehicles(UAVs).UAVs can fly in high altitude and collect marine environment information on ***,UAVs can plan defense paths for USVs to intercept intruders with full-assignment or reassignment strategies aiming at maximum overall ***,we propose dynamic overlay reconnaissance algor计hm based on genetic idea(GI-DORA)to solve the problem of multi-UAV multi-station ***,we develop continuous particle swarm optimization based on obstaele dimension(OD-CPSO)to optimize defense path of USVs to intercept *** addition,under the designed defense constraints,we propose dispersed particle swarm optimization based on mutation and crossover(MC-DPSO)and real-time batch assignment algorithm(RTBA)in emergency for formulating combat defense mission assignment strategy in different ***,we illus trate the feasibility and effectiveness of the proposed met hods.

关键词： collaborative defense mission assignment path planning unmanned surface vehicles(USVs) unmanned aerial vehicles(UAVs)

来源：评论

学校读者我要写书评

暂无评论

UniEmoX: Cross-modal Semantic-Guided Large-Scale Pretraining for Universal Scene Emotion Perception

arXiv

引用

arXiv 2024年

作者： Chen, Chuang Sun, Xiao Liu, Zhi School of Artificial Intelligence Anhui University Hefei230601 China Institute of Artificial Intelligence Hefei Comprehensive National Science Center Hefei230088 China Anhui Province Key Laboratory of Affective Computing and Advanced Intelligent Machines School of Computer Science and Information Engineering Hefei University of Technology Hefei230009 China Department of Computer and Network Engineering The University of Electro-Communications Chofu-shi Tokyo182-8585 Japan

Visual emotion analysis holds significant research value in both computer vision and psychology. However, existing methods for visual emotion analysis suffer from limited generalizability due to the ambiguity of emotion perception and the diversity of data scenarios. To tackle this issue, we introduce UniEmoX, a cross-modal semantic-guided large-scale pretraining framework. Inspired by psychological research emphasizing the inseparability of the emotional exploration process from the interaction between individuals and their environment, UniEmoX integrates scene-centric and person-centric low-level image spatial structural information, aiming to derive more nuanced and discriminative emotional representations. By exploiting the similarity between paired and unpaired image-text samples, UniEmoX distills rich semantic knowledge from the CLIP model to enhance emotional embedding representations more effectively. To the best of our knowledge, this is the first large-scale pretraining framework that integrates psychological theories with contemporary contrastive learning and masked image modeling techniques for emotion analysis across diverse scenarios. Additionally, we develop a visual emotional dataset titled Emo8. Emo8 samples cover a range of domains, including cartoon, natural, realistic, science fiction and advertising cover styles, covering nearly all common emotional scenes. Comprehensive experiments conducted on six benchmark datasets across two downstream tasks validate the effectiveness of UniEmoX. The source code is available at https://***/chincharles/u-emo. Copyright © 2024, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

A Privacy Preserving and Byzantine Robust Collaborative Federated Learning Method Design

A Privacy Preserving and Byzantine Robust Collaborative Fede...

引用

IEEE International Conference on Communications (ICC)

作者： Nuocheng Yang Sihua Wang Mingzhe Chen Changchuan Yin Christopher G. Brinton Beijing Laboratory of Advanced Information Network Beijing University of Posts and Telecommunications Beijing China State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China Department of Electrical and Computer Engineering Institute for Data Science and Computing University of Miami Coral Gables FL USA School of Electrical and Computer Engineering Purdue University West Lafayette IN USA

ISBN: (数字)9781728190549

ISBN: (纸本)9781728190556

Collaborative federated learning (CFL) enables device cooperation in training shared machine learning models without reliance on a parameter server. However, the absence of a parameter server also impacts vulnerabilities associated with adversarial attacks, including privacy inference and Byzantine attacks. In this context, this paper introduces a novel CFL framework that enables each device to individually determine the subset of devices to transmit FL parameters to over the wireless network, based on its neighboring devices' location, current loss, and connection information, to achieve privacy protection and robust aggregation. This is formulated as an optimization problem whose goal is to minimize CFL training loss while satisfying the privacy preservation, robust aggregation, and transmission delay requirements. To solve this problem, a proximal policy optimization (PPO)-based reinforcement learning (RL) algorithm integrated with a graph neural network (GNN) is proposed. Compared to traditional algorithms that use global information with high computational complexity, the proposed GNN-RL method can be deployed on devices based on neighboring information with lower computational overhead. Simulation results show that the proposed algorithm can protect data privacy and increase identification accuracy by 15% compared to an algorithm in which devices are partially clustered for model aggregation.

关键词： Training Privacy Federated learning Simulation Collaboration Clustering algorithms Propagation losses

来源：评论

学校读者我要写书评

暂无评论

WaveFormer: A Wavelet Transformer for Parkinson Disease’s Retinal Layer Segmentation in OCT

WaveFormer: A Wavelet Transformer for Parkinson Disease’s R...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Yanlin Chen Xiaoqing Zhang Tianao Wang Chen Tang Haili Ye Jiang Liu Department of Computer Science and Engineering Research Institute of Trustworthy Autonomous Systems Southern University of Science and Technology Shenzhen China Shenzhen Institute of Advanced Technology Center for High Performance Computing and Shenzhen Key Laboratory of Intelligent Bioinformatics Chinese Academy of Sciences Shenzhen China School of Ophthalmology and Optometry Wenzhou Medical University Wenzhou China The Oujiang Laboratory The Affiliated Eye Hospital Wenzhou Medical University Wenzhou China

ISBN: (数字)9798350359312

ISBN: (纸本)9798350359329

Pathology symptoms of Parkinson disease (PD) are different from those of retinal diseases in the retinal layers, which are subtle. However, segmenting pathology information of PD from retinal layers automatically based on optical coherence tomography (OCT) images has not been studied before. Although existing Transformer-based segmentation methods have achieved good segmentation results, they have limitations in capturing local context information. Convolutional neural networks (CNNs) can construct local context dependencies among pixels, which is complementary to Transformers. Particularly, edge information extraction is significant for accurate retinal layer segmentation, which is ignored by both Transformers and CNNs but can be captured by frequency domain learning methods. To fully leverage the advantages of Transformers, CNNs, and frequency domain learning methods, we propose a Wavelet Transformer (WaveFormer) for retinal layer segmentation based on OCT images. In the WaveFormer, we design a Wavelet Spatial Attention block to exploit the potential of frequency information. Based on these advantages, WaveFormer can be data-efficient in limited OCT images of PD. The experimental results on the OCT-PD segmentation dataset show that our WaveFormer outperforms existing Transformers and CNNs. For example, WaveFormer outperforms Swin-UNet by 3.41% of IoU.

关键词： Learning systems Image segmentation Pathology Wavelet domain Image edge detection Frequency-domain analysis Transformers

来源：评论

学校读者我要写书评

暂无评论

A Machine Learning-based Automatic Model to Predicting Performance of Students

A Machine Learning-based Automatic Model to Predicting Perfo...

引用

Current Development in Engineering and Technology (CCET), IEEE International Conference on

作者： Atul Garg Nidhi Bansal Garg Pinaki Ghosh Ankit Bansal Umesh Kumar Lilhore Sarita Simaiya Chitkara University Institute of Engineering and Technology Chitkara University Punjab India School of Advanced Computing SAGE University Bhopal India Department of Computer Science and Engineering Chandigarh University Mohali Punjab India

Machine learning offers a variety of techniques for use in diverse fields. Massive quantities of student data are produced in educational institutions, and machine-learning techniques are a precious tool for identifying patterns in individuals' behavior. Academicians can learn valuable data concerning the students' expertise and how it relates to their educational assignments by critically examining and evaluating these data. This data serves as the input for cutting-edge techniques and algorithms that can forecast students' academic achievement. This research presents a machine learning-based automatic model for predicting the performance of students. The popular ML methods are Random Forest, K-Neighbors, Extra Trees, Ada-Boost, Decision Tree, Logistic Regression, Gaussian NB, and Bernoulli NB. This research utilizes an online student performance dataset collected from Kaggle. The findings demonstrate that Random forest and decision trees outperform other techniques based on high precision as well as the capacity to interpret the guidelines and structure that were developed for predicting students' educational success.

关键词： Machine learning algorithms Automation Education Forestry Predictive models Prediction algorithms Behavioral sciences

来源：评论

学校读者我要写书评

暂无评论

Towards a Sustainable Future: Analyzing Water Quality Changes for Environmental Sustainability

Towards a Sustainable Future: Analyzing Water Quality Change...

引用

Smart Technology (ICE-SMARTec), IEEE International Conference on

作者： Nur Suhaili Mansor Hapini Awang Nor Hazlyna Harun Mohamad Ghozali Hassan Sarkin Tudu Shehu Malami Noraishah Shamsuddin Institute for Advanced and Smart Digital Opportunities School of Computing Universiti Utara Malaysia Sintok Malaysia Data Science Research Lab School of Computing Universiti Utara Malaysia Sintok Malaysia Disaster Management Institute (DMI) School of Technology Management and Logistics Universiti Utara Malaysia Sintok Malaysia Department of Computer Science and Information Technology Faculty of Science Sokoto State University Sokoto Nigeria ICI in Smart Manufacturing Sirim Bhd Shah Alam Malaysia

The development of industrial areas is frequently thought to have a good impact on society because it helps to meet the population's requirements. However, it also negatively impacts the environment, such as river pollution. If a country experiences rapid economic growth without considering the effects of industrial development on the environment, then human well-being will be affected. In this context, advanced development is attained by a developed nation while ensuring the well-being of society. Uncontrolled and unsustainable development increases environmental pollution, adversely affecting human health. As water is essential for human life, water resources, such as rivers, must be properly managed and protected to ensure human well-being. Thus, the study conducted on the changes in the water quality of the Kim Kim River provides insights into how unintended consequences of urban and industrial development can be identified and addressed using three sets of satellite imagery data from 2015, 2019 and 2020. The study was led to identify research problems and achieve the study objectives to obtain research results. The results of image classification analyses indicate that the development area near Sungai Kim Kim increased to 50%, with 38% from industrial areas. The RMSE for the image classification is 1.51, significantly better than the minimum allowed value of 2.00. It provides an understanding of the change in river water quality because of urban and industrial development. By highlighting the importance of sustainable development and the need to balance economic growth with environmental protection, this study contributes to the development of effective policies and strategies for the sustainable development of industrial areas.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Training-free image style alignment for self-adapting domain shift on handheld ultrasound devices

arXiv

引用

arXiv 2024年

作者： Zeng, Hongye Zou, Ke Chen, Zhihao Gao, Yuchong Chen, Hongbo Zhang, Haibin Zhou, Kang Wang, Meng Goh, Rick Siow Mong Liu, Yong Jiang, Chang Zheng, Rui Fu, Huazhu School of Information Science and Technology ShanghaiTech University Shanghai China Shanghai Advanced Research Institute Chinese Academy of Sciences Shanghai China University of Chinese Academy of Sciences Beijing China National Key Laboratory of Fundamental Science on Synthetic Vision College of Computer Science Sichuan University Chengdu China College of Intelligence and Computing Tianjin University Tianjin China Department of Computer Science and Engineering The Chinese University of Hong Kong Hong Kong Beth Israel Deaconess Medical Center Harvard Medical School Boston United States Institute of High Performance Computing Agency for Science Technology and Research Singapore Singapore Department of Orthopedics Zhongshan Hospital Fudan University Shanghai China Shanghai Engineering Research Center of Intelligent Vision and Imaging Shanghai China

Handheld ultrasound devices face usage limitations due to user inexperience and cannot benefit from supervised deep learning without extensive expert annotations. Moreover, the models trained on standard ultrasound device data are constrained by training data distribution and perform poorly when directly applied to handheld device data. In this study, we propose the Training-free Image Style Alignment (TISA) framework to align the style of handheld device data to those of standard devices. The proposed TISA can directly infer handheld device images without extra training and is suited for clinical applications. We show that TISA performs better and more stably in medical detection and segmentation tasks for handheld device data. We further validate TISA as the clinical model for automatic measurements of spinal curvature and carotid intima-media thickness. The automatic measurements agree well with manual measurements made by human experts and the measurement errors remain within clinically acceptable ranges. We demonstrate the potential for TISA to facilitate automatic diagnosis on handheld ultrasound devices and expedite their eventual widespread use. © 2024, CC BY-NC-SA.

关键词： Deep learning

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：