检索结果-内蒙古大学图书馆

Classification of Breast Thermal Images into Healthy/Cancer Group Using Pre-Trained Deep Learning Schemes

Procedia Computer Science 2023年 218卷 24-34页

作者： Seifedine Kadry Rubén González Crespo Enrique Herrera-Viedma Sujatha Krishnamoorthy Venkatesan Rajinikanth Faculty of Applied Computing and Technology Noroff University College Kristiansand 94612 Norway Computer Science Department School of Engineering and Technology Universidad Internacional de La Rioja 26006 Andalusian Spain Research Institute in Data Science and Computational Intelligence University of Granada Granada Spain Zhejiang Bioinformatics International Science and Technology Cooperation Center Wenzhou-Kean University Zhejiang Province China Wenzhou Municipal Key Lab of Applied Biomedical and Biopharmaceutical Informatics Wenzhou-Kean University Zhejiang Province China Department of Computer Science and Engineering Saveetha School of Engineering SIMATS Chennai 602105 Tamil Nadu India

In the women's community, Breast Cancer (BC) is a severe disease. The World Health Organization reported in 2020 that 2.26 million deaths occur due to BC. BC is curable if detected early. Since thermal imaging is non-invasive and supports disease detection, it is commonly used in clinics. Compared to other methods, it keeps BC early and accurate. The proposed work aims to evaluate the performance of the Pretrained Deep-Learning Methods (PDLM) in detecting BC using the thermal images collected from the benchmark dataset. It includes the following stages: primary image processing, deep feature mining, handcrafted feature mining, feature optimization using Firefly-Algorithm (FA), classification and validation. Visual lab thermal images were used in the study. The investigational outcome of this study authenticates that the VGG16, along with the DT, provides better detection accuracy (95.5%) compared to other classifiers used in this study. To justify the significance of the implemented technique, the proposed work not only improved accuracy, but also improved precision, sensitivity, specificity, and F1-Scores.

关键词： Breast cancer Thermal imaging Firefly algorithm Feature optimization Classification

来源：评论

学校读者我要写书评

暂无评论

Divert More Attention to Vision-Language Tracking

arXiv

引用

arXiv 2022年

作者： Guo, Mingzhe Zhang, Zhipeng Fan, Heng Jing, Liping Beijing Key Lab of Traffic Data Analysis and Mining Beijing Jiaotong University China Nlpr Institute of Automation Chinese Academy of Sciences China Department of Computer Science and Engineering University of North Texas United States

Relying on Transformer for complex visual feature learning, object tracking has witnessed the new standard for state-of-the-arts (SOTAs). However, this advancement accompanies by larger training data and longer training period, making tracking increasingly expensive. In this paper, we demonstrate that the Transformer-reliance is not necessary and the pure ConvNets are still competitive and even better yet more economical and friendly in achieving SOTA tracking. Our solution is to unleash the power of multimodal vision-language (VL) tracking, simply using ConvNets. The essence lies in learning novel unified-adaptive VL representations with our modality mixer (ModaMixer) and asymmetrical ConvNet search. We show that our unified-adaptive VL representation, learned purely with the ConvNets, is a simple yet strong alternative to Transformer visual features, by unbelievably improving a CNN-based Siamese tracker by 14.5% in SUC on challenging LaSOT (50.7%→65.2%), even outperforming several Transformer-based SOTA trackers. Besides empirical results, we theoretically analyze our approach to evidence its effectiveness. By revealing the potential of VL representation, we expect the community to divert more attention to VL tracking and hope to open more possibilities for future tracking beyond Transformer. Code and models will be released at https://***/JudasDie/SOTS. Copyright © 2022, The Authors. All rights reserved.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Secure Deduplication Against Frequency Analysis Attacks

Secure Deduplication Against Frequency Analysis Attacks

引用

International Conference on Mobile Ad-hoc and Sensor Networks, MSN

作者： Hang Chen Guanxiong Ha Yuchen Chen Haoyu Ma Chunfu Jia College of Cyber Science Nankai University Tianjin China Tianjin Key Laboratory of Network and Data Security Technology Tianjin China Zhejiang Lab Hangzhou China The School of Cyber Engineering Xidian University Xi’an China

Message-locked Encryption (MLE) is the most common approach used in encrypted deduplication systems. However, the systems based on MLE are vulnerable to frequency analysis attacks, because MLE encrypts the identical plain texts into the identical ciphertexts, which is deterministic. The state-of-the- art defense scheme, which named TED, lacks key verification and uses a single key server to record frequency information. Once the key server is compromised, TED will be vulnerable to brute-force attacks. In addition, TED's key generation algorithm needs to be designed more exquisitely to strengthen protection, and its security indicator is not comprehensive. We propose SDAF, which supports key verification and enhanced protection against frequency analysis attacks. Based on chameleon hash, SDAF realizes key verification to prevent malicious key servers from generating fake encryption keys. In order to disturb the frequency information, SDAF introduces reservoir sample to generate uniformly distributed encryption keys, and uses multiple key servers, which interact with each other via multi-party PSI and rotate spontaneously to avoid the single point of failure. Moreover, a new indicator Kurtosis is pointed out to evaluate the security against frequency analysis attacks. We implement the prototypes of SDAF. The experiments of the real-world data sets show that, compared with the existing schemes, SDAF can better resist frequency analysis attacks with lower time overheads.

关键词： Maximum likelihood estimation Time-frequency analysis data privacy Art Prototypes Resists Reservoirs

来源：评论

学校读者我要写书评

暂无评论

Collaborative Edge Service Placement for Maximizing QoS with Distributed data Cleaning 31

Collaborative Edge Service Placement for Maximizing QoS with...

引用

31st IEEE/ACM International Symposium on Quality of Service, IWQoS 2023

作者： Liang, Yuzhu Wang, Wenhua Zheng, Xi Liu, Qin Wang, Liang Wang, Tian School of Artificial Intelligence Beijing Normal University Beijing100875 China School of Computing Macquarie University NSW2109 Australia College of Computer Science and Electronic Engineering Hunan University Changsha410082 China School of Computer Science Northwestern Polytechnical University Xi'an710072 China BNU-UIC Institute of Artificial Intelligence and Future Networks Beijing Normal University Zhuhai519000 China BNU-HKBU United International College Guangdong Key Lab of Ai and Multi-Modal Data Processing Zhuhai519000 China

ISBN: (纸本)9798350399738

The proliferation of dirty data on Internet of Things (IoT) devices can undermine the accuracy of data-driven decision-making by affecting the distribution of original data. The Quality of Service (QoS) of data cleaning on these devices is heavily impacted by processing delay and accuracy. In this paper, we find that edge service placement is a key step aligned with data cleaning and consider the collaborative edge service placement with distributed data cleaning (SPDC) problem. To address this issue, we propose a novel distributed collaborative edge-based architecture that effectively balances the demands of storage, communication, computation, and load constraints. Experimental results show that the proposed approach significantly improves the accuracy of data cleaning by 0.31%-86.07% and reduces delay by 2.73%-58.71% compared to state-of-the-art baselines. © 2023 IEEE.

关键词： Edge computing

来源：评论

学校读者我要写书评

暂无评论

Improved Visual Fine-tuning with Natural Language Supervision

Improved Visual Fine-tuning with Natural Language Supervisio...

引用

International Conference on Computer Vision (ICCV)

作者： Junyang Wang Yuanhong Xu Juhua Hu Ming Yan Jitao Sang Qi Qian School of Computer and Information Technology & Beijing Key Lab of Traffic Data Analysis and Mining Beijing Jiaotong University Beijing China DAMO Academy Alibaba Group Hangzhou China School of Engineering and Technology University of Washington Tacoma WA USA Peng Cheng Lab Shenzhen China DAMO Academy Alibaba Group Bellevue WA USA

Fine-tuning a visual pre-trained model can leverage the semantic information from large-scale pre-training data and mitigate the over-fitting problem on downstream vision tasks with limited training examples. While the problem of catastrophic forgetting in pre-trained backbone has been extensively studied for fine-tuning, its potential bias from the corresponding pre-training task and data, attracts less attention. In this work, we investigate this problem by demonstrating that the obtained classifier after fine-tuning will be close to that induced by the pre-trained model. To reduce the bias in the classifier effectively, we introduce a reference distribution obtained from a fixed text classifier, which can help regularize the learned vision classifier. The proposed method, Text Supervised fine-tuning (TeS), is evaluated with diverse pre-trained vision models including ResNet and ViT, and text encoders including BERT and CLIP, on 11 downstream tasks. The consistent improvement with a clear margin over distinct scenarios confirms the effectiveness of our proposal. Code is available at https://***/idstcv/TeS.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Robust zero-watermarking algorithm for medical images based on Hadamard-DWT-DCT

引用

International Journal of Wireless and Mobile Computing 2022年第2期23卷 183-192页

作者： Sheng, Mingshuai Li, Jingbing Liu, Jing School of Information and Communication Engineering Hainan University Hainan Haikou China State Key Laboratory of Marine Resource Utilisation in the South China Sea Hainan University Hainan Haikou China Research Centre for Healthcare Data Science Zhejiang Lab Hangzhou China

As the IT industry grows rapidly, information security is particularly important. Digital development is a double-edged sword for the medical field, which not only brings convenience to patients and doctors, but also has hidden dangers for information security. We propose a Hadamard-DWT-DCT-based zero watermark algorithm for medical images for the problem of privacy information leakage when medical images are transmitted on the internet. First, the raw medical image is chunked using the Hadamard transformation and produces a coefficient matrix, which is then transformed into a wavelet coefficient matrix, which can be efficiently compressed and stored. The wavelet coefficient matrix is then DCT-transformed to finally obtain the eigenvector. Experimental results show that the proposed algorithm is not only highly robust, and the watermark can still be effectively extracted against different degrees of geometric and conventional attack interference, but also can encrypt the patient privacy information contained in the medical image. Copyright © 2022 Inderscience Enterprises Ltd.

关键词： Medical imaging

来源：评论

学校读者我要写书评

暂无评论

Generalization of CNNs on Relational Reasoning with Bar Charts

arXiv

引用

arXiv 2025年

作者： Cui, Zhenxing Chen, Lu Wang, Yunhai Haehn, Daniel Wang, Yong Pfister, Hanspeter School of Computer Science and Technology Shandong University China State Key Lab of CAD&CG Zhejiang University China School of Information Renmin University of China China College of Science and Mathematics University of Massachusetts Boston United States College of Computing and Data Science Nanyang Technological University Singapore School of Engineering and Applied Sciences Harvard University United States

This paper presents a systematic study of the generalization of convolutional neural networks (CNNs) and humans on relational reasoning tasks with bar charts. We first revisit previous experiments on graphical perception and update the benchmark performance of CNNs. We then test the generalization performance of CNNs on a classic relational reasoning task: estimating bar length ratios in a bar chart, by progressively perturbing the standard visualizations. We further conduct a user study to compare the performance of CNNs and humans. Our results show that CNNs outperform humans only when the training and test data have the same visual encodings. Otherwise, they may perform worse. We also find that CNNs are sensitive to perturbations in various visual encodings, regardless of their relevance to the target bars. Yet, humans are mainly influenced by bar lengths. Our study suggests that robust relational reasoning with visualizations is challenging for CNNs. Improving CNNs’ generalization performance may require training them to better recognize task-related visual properties. Copyright © 2025, The Authors. All rights reserved.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Seasonal Variation, Gas/Particle Partitioning, Sources and Risks of Polycyclic Aromatic Compounds in Xi’an, Northwest China

SSRN

引用

SSRN 2023年

作者： Zhang, Jianing Wei, Chong Han, Yongming Lei, Dewen State Key Laboratory of Loess and Quaternary Geology Key Lab of Aerosol Chemistry and Physics Institute of Earth Environment Chinese Academy of Sciences Xi’an710061 China Shanghai Carbon Data Research Center CAS Key Laboratory of Low-Carbon Conversion Science and Engineering Shanghai Advanced Research Institute Chinese Academy of Sciences Shanghai201210 China Department of Environmental Science and Technology School of Human Settlements and Civil Engineering Xi’an Jiaotong University Xi’an710049 China University of Chinese Academy of Sciences Beijing100049 China

Polycyclic aromatic compounds (PACs), such as polycyclic aromatic hydrocarbons (PAHs,) and their derivatives (oxygenated PAHs, nitrated PAHs and N-heterocyclic PAHs), have received extensive attention due to their carcinogenicity, teratogenicity, and mutagenicity to humans. In this study, gaseous and PM2.5-bound PACs were analysed in four seasons in 2016 in Xi’an and further evaluated for their gas/particle partition, sources, and health risks. Annual average concentrations (gaseous + particulate) of Σ27PAHs, Σ12OPAHs, Σ14NPAHs, and Σ4AZAs were 445.1±428.0, 45.6±62.4, 4.9±6.5, and 6.2±8.6 ng m−3, respectively, with the highest values in winter and the lowest values in summer. Low-molecular-weight PACs mainly occur in gaseous and high-molecular-weight PACs partition in particles, and the major mechanism is absorption into the atmospheric particles. PACs significantly correlated with combustion-derived air pollutants and carbon fractions, suggesting their common sources and transport pathways. The concentration ratios indicated mixed sources of fossil fuel combustion and biomass burning, while the positive matrix factorization model suggested that coal combustion, vehicle exhaust, and biomass burning accounted for 30.0%, 25.8% and 23.6% of PAHs, respectively, and direct emissions contributed 56.1% of OPAHs. The PACs estimated annual lifetime lung cancer risk is 2.67 × 10−5, exceeding the value advocated by the World Health Organization (10−5). © 2023, The Authors. All rights reserved.

关键词： Polycyclic aromatic hydrocarbons

来源：评论

学校读者我要写书评

暂无评论

Incremental Learning Algorithm Based on Graph Regularized Non-negative Matrix Factorization with Sparseness Constraints 4

Incremental Learning Algorithm Based on Graph Regularized No...

引用

4th International Conference on Artificial Intelligence and Big data, ICAIBD 2021

作者： Wang, Jintao Zhang, Meng Hu, Xusheng Ni, Tianwei Key Lab. of Unmanned Aer. Vehicle Devmt. and Data Application of Anhui Higher Education Institutes Maanshan Wireless Sensor Network and Intelligent Perception Engineering Technology Research Center Wanjiang University of Technology Maanshan China Wanjiang University of Technology Maanshan China

ISBN: (纸本)9780738131702

Focused on the issues that the sparseness of the data obtained after factorization is reduced and with the increasing of training samples, the computing scale increases rapidly, an incremental learning algorithm based on graph regularized non-negative matrix factorization with sparseness constraints was proposed. It not only considers the geometric structure in the data representation, but also introduces sparseness constraint to coefficient matrix and combines them with incremental learning. Using the results of previous factorization involved in iterative computation with sparseness constraints and graph regularized, the cost of the computation is reduced and the sparseness of data after factorization is highly improved. Experiments on both ORL and PIE face recognition databases demonstrate the effectiveness of the proposed method. © 2021 IEEE.

关键词： Non-negative matrix factorization

来源：评论

学校读者我要写书评

暂无评论

PRAD: Periapical Radiograph Analysis dataset and Benchmark Model Development

arXiv

引用

arXiv 2025年

作者： Zhou, Zhenhuan Zhang, Yuchen Xu, Ruihong Zhao, Xuansen Li, Tao College of Computer Science Nankai University Tianjin300350 China Key Laboratory of Data and Intelligent System Security Ministry of Education China Department of stomatology Tianjin Union Medical Center The First Affiliated Hospital of Nankai University Tianjin300121 China College of Mechanical Engineering Taiyuan University of Technology Shanxi 030000 China Haihe Lab of ITAI Tianjin300459 China

Deep learning (DL), a pivotal technology in artificial intelligence, has recently gained substantial traction in the domain of dental auxiliary diagnosis. However, its application has predominantly been confined to imaging modalities such as panoramic radiographs and Cone Beam Computed Tomography, with limited focus on auxiliary analysis specifically targeting Periapical Radiographs (PR). PR are the most extensively utilized imaging modality in endodontics and periodontics due to their capability to capture detailed local lesions at a low cost. Nevertheless, challenges such as resolution limitations and artifacts complicate the annotation and recognition of PR, leading to a scarcity of publicly available, large-scale, high-quality PR analysis datasets. This scarcity has somewhat impeded the advancement of DL applications in PR analysis. In this paper, we present PRAD-10K, a dataset for PR analysis. PRAD-10K comprises 10,000 clinical periapical radiograph images, with pixel-level annotations provided by professional dentists for nine distinct anatomical structures, lesions, and artificial restorations or medical devices, We also include classification labels for images with typical conditions or lesions. Furthermore, we introduce a DL network named PRNet to establish benchmarks for PR segmentation tasks. Experimental results demonstrate that PRNet surpasses previous state-of-the-art medical image segmentation models on the PRAD-10K dataset. The codes and dataset will be made publicly available. © 2025, CC BY-NC-ND.

关键词： Dental orthoses

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：