检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Gong, Haifan Huang, Wenhao Zhang, Huan Wang, Yu Wan, Xiang Shen, Hong Li, Guanbin Li, Haofeng Shenzhen Research Institute of Big Data China InferVision China School of Computer Science and Engineering Sun Yat-sen University China School of Science and Engineering The Chinese University of Hong Kong Shenzhen China

Automatic segmentation of the bronchial tree from CT imaging is important, as it provides structural information for disease diagnosis. Despite the merits of previous automatic bronchus segmentation methods, they have paied less attention to the issue we term as Intensity Confusion, wherein the intensity values of certain background voxels approach those of the foreground voxels within bronchi. Conversely, the intensity values of some foreground voxels are nearly identical to those of background voxels. This proximity in intensity values introduces significant challenges to neural network methodologies. To address the issue, we introduce a novel Intensity-Distance Guided loss function, which assigns adaptive weights to different image voxels for mining hard samples that cause the intensity confusion. The proposed loss estimates the voxel-level hardness of samples, on the basis of the following intensity and distance priors. We regard a voxel as a hard sample if it is in: (1) the background and has an intensity value close to the bronchus region;(2) the bronchus region and is of higher intensity than most voxels inside the bronchus;(3) the background region and at a short distance from the bronchus. Extensive experiments not only show the superiority of our method compared with the state-of-the-art methods, but also verify that tackling the intensity confusion issue helps to significantly improve bronchus segmentation. Project page: https://***/lhaof/ICM. Copyright © 2024, The Authors. All rights reserved.

关键词： computerized tomography

来源：评论

学校读者我要写书评

暂无评论

A Multimodal Fake News Detection Model Based on Cross-Image Semantic Fusion

A Multimodal Fake News Detection Model Based on Cross-Image ...

引用

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

作者： Huanran Wang Yongxin Yang Shuai Han Zhenyuan He Wu Yang CETC Big Data Research Institute Co. Ltd China School of Computer Science and Technology Harbin Engineering University China

ISBN: (数字)9798331516024

ISBN: (纸本)9798331516031

In recent years, social media has become one of the most popular ways of news dissemination. There is a phenomenon that numerous fake news are spreading rampantly on public social media platforms, posing a serious threat to the credibility of social media. Moreover, more and more social media news posts carry multimodal contexts, i.e., utilize not only text but also abundant images to describe the news. However, existing methods only involve the first image along with text in multimodal fake news detection. It severely hampers the extraction of global image semantic information and consequently damages the effectiveness of multimodal fake news detection. To address this issue, we propose a Cross-Image Semantic Fusion based multi-modal fake news detection method (CISF for short). The method uses an adaptive attention diffusion module to model semantic correlations among different images, fully leveraging the contextual dependencies between different images to achieve semantic interaction and fusion among images. On the basis, a global image semantic representation is generated to represent the entire image modality. Finally, the fake news detection is performed based on the multimodal fusion of the text and the global image semantic representation. We conduct experiments on two real-world datasets and demonstrate the effectiveness of the proposed method.

关键词： Adaptation models Limiting Correlation Social networking (online) Fuses Semantics Sensors data mining Fake news Context modeling

来源：评论

学校读者我要写书评

暂无评论

Triple-negative Breast Cancer Classification Algorithm based on High-dimensional Big data of Breast Ultrasound Radiomics

引用

Journal of engineering science and Technology Review 2023年第1期16卷 85-91页

作者： Liu, Wen Yao, Hui Leng, Xiaoling Xue, Yutong Ma, Binlin College of Computer Science and Technology Xinjiang Normal University Urumqi830054 China Artificial Intelligence and Smart Mine Engineering Technology Center Xinjiang Institute of Engineering Urumqi830023 China Xinjiang Changsen Data Technology Co. Ltd Urumqi830011 China Department of Ultrasound The Affiliated Tumor Hospital of Xinjiang Medical University Urumqi830011 China RestonVA20191 United States

In computer-aided diagnosis, breast cancer classification accuracy of artificial intelligence (AI) method might be influenced by imbalanced classification samples. Model training efficiency also decreases with the increase in data size. To decrease the influences of sample classification imbalance and increase data size on breast cancer classification efficiency, this study proposed a triple-negative breast cancer classification algorithm (TNBCC) based on similarity query. Similar breast cancer feature data could be stored in a similar region by establishing the TNBCC index. A pruning algorithm was designed based on the TNBCC index, which decreases similarity measurement with nodes of the index tree significantly during data input and query. Users could find data similar to the query object quickly and accurately through the TNBCC index using pruning algorithm, and they could obtain the molecular subtype of the query object according to the searched data. Results demonstrate that, on the real dataset provided by the Affiliated Tumor Hospital of Xinjiang Medical University, the TNBCC algorithm decreases the influences of sample classification imbalance on accuracy effectively compared with the existing AI algorithm, and the classification accuracy reaches 91%. Given the mass high-dimensional data, it takes 49 s to build the TNBCC index tree, with an efficiency higher than that of the existing AI algorithm. This study provides a good reference to improve the performances of breast cancer classification algorithm. © 2023 School of science, IHU. All rights reserved.

关键词： computer aided diagnosis

来源：评论

学校读者我要写书评

暂无评论

Estimating Physical Information Consistency of Channel data Augmentation for Remote Sensing Images

Estimating Physical Information Consistency of Channel Data ...

引用

IEEE International Symposium on Geoscience and Remote Sensing (IGARSS)

作者： Tom Burgert Begüm Demir BIFOLD – Berlin Institute for the Foundations of Learning and Data Germany Faculty of Electrical Engineering and Computer Science Technische Universität Berlin Germany

ISBN: (数字)9798350360325

ISBN: (纸本)9798350360332

The application of data augmentation for deep learning (DL) methods plays an important role in achieving state-of-the-art results in supervised, semi-supervised, and self-supervised image classification. In particular, channel transformations (e.g., solarize, grayscale, brightness adjustments) are integrated into data augmentation pipelines for remote sensing (RS) image classification tasks. However, contradicting beliefs exist about their proper applications to RS images. A common point of critique is that the application of channel augmentation techniques may lead to physically inconsistent spectral data (i.e., pixel signatures). To shed light on the open debate, we propose an approach to estimate whether a channel augmentation technique affects the physical information of RS images. To this end, the proposed approach estimates a score that measures the alignment of a pixel signature within a time series that can be naturally subject to deviations caused by factors such as acquisition conditions or phenological states of vegetation. We compare the scores associated with original and augmented pixel signatures to evaluate the physical consistency. Experimental results on a multi-label image classification task show that channel augmentations yielding a score that exceeds the expected deviation of original pixel signatures can not improve the performance of a baseline model trained without augmentation.

关键词： Time series analysis Pipelines Channel estimation Vegetation mapping data augmentation Time measurement Sensors

来源：评论

学校读者我要写书评

暂无评论

A Model Based on Radiomics and Machine Learning in Glioma Grading 2022

A Model Based on Radiomics and Machine Learning in Glioma Gr...

引用

2nd International Conference on Bioinformatics and Intelligent Computing, BIC 2022

作者： Wang, Junxiu Zeng, Jianchao Yu, Xiaoqing Liu, Jingang Data Science and Technology North University of China Shanxi Province Taiyuan030051 China Department of Computer Engineering Taiyuan Institute of Technology Shanxi Taiyuan030008 China

ISBN: (纸本)9781450395755

Radiomics-based researches have shown the predictive abilities of machine learning methods in medical diagnosis. However, different machine learning approaches affect the prediction performance. This paper proposes a method based on Tree-based Pipeline Optimization Tool (TPOT) to find the best classification method in glioma grading. This study utilized the public multimodal Brain Tumor Segmentation Challenge (BraTS) 2019 magnetic resonance imaging (MRI) dataset. 3860 radiomics features were extracted from multi-modal MRI images, including tumor morphological features, first-order gray features, texture features, etc. Then the least absolute shrinkage and selection operator (LASSO) was used to select 88 best radiomics features. Finally, the TPOT was used to construct the brain glioma grade prediction model based on the selected features. The accuracy of the model optimized by TPOT was 100% and the area under the ROC )AUC( was 1 in the training group, and 95.52% and 0.98 in the test group, respectively. Based on machine learning algorithms, brain glioma can be graded automatically by radiomics method. © 2022 ACM.

关键词： Grading

来源：评论

学校读者我要写书评

暂无评论

Revolutionizing COVID-19 Diagnosis with Radiographic Images and Deep Learning: A Paradigm Shift in Healthcare

Revolutionizing COVID-19 Diagnosis with Radiographic Images ...

引用

Next Generation Electronics (NEleX), International Conference on

作者： Mohanaprakash T A S. Saranya M. Abinaya K. C. Rajheshwari D. Elangovan M. Gobinath Department of Computer Science and Engineering Panimalar Engineeering College Chennai India Department of Artificial Intelligence and Data Science Rajalakshmi Institute of Technology Chennai India Department of Computer Science and Engineering Sri Eshwar College of Engineering Coimbatore India Department of Computer Science and Engineering Sona College of Technology Salem TamilNadu Departement of Computer and Communication Engineering Sri Manakula Vinayagar Engineering College Puducherry

Amid the relentless battle against the COVID-19 pandemic, a cadre of researchers and scientists is diligently crafting innovative tools to expedite diagnosis and treatment. One standout innovation is the integration of radiographic images and deep convolutional neuronal networks (CNN), offering deeper insights into COVID-19. Early, accurate diagnosis is paramount in managing the COVID-19 crisis. It not only facilitates timely treatment but also curbs the virus's spread. The amalgamation of radiographic images and advanced machine learning techniques emerges as a transformative force. Traditional COVID-19 diagnostics rely heavily on molecular tests, like Polymerase Chain Reaction (PCR) tests. While highly accurate, these tests often involve protracted processing times and substantial resource demands. They may also produce false negatives in early infections. Radiographic imaging, especially chest X-rays, addresses these limitations. Researchers have harnessed deep convolutional neuronal networks, specialized for image analysis, to create an innovative diagnostic platform. This platform meticulously scrutinizes chest X-ray images, distinguishing COVID-19 pneumonia from other forms with exceptional precision. The deep CNN model undergoes intensive training on a diverse dataset, including chest X-rays from both COVID-19-positive and COVID-19-negative patients. It employs intricate processes of feature extraction and pattern recognition, discerning subtle nuances indicative of COVID-19-related lung abnormalities. This proficiency empowers the model to categorize new chest X-ray images, assisting radiologists in decision-making. The merits of this groundbreaking approach are multifaceted. Foremost is the substantial reduction in diagnosis time. Unlike molecular tests, which can take hours, chest X-ray analysis yields results within minutes, facilitating swift patient triage. It holds particular promise in resource-constrained settings where molecular testing may be scarc

关键词：

来源：评论

学校读者我要写书评

暂无评论

HYCOT: A TRANSFORMER-BASED AUTOENCODER FOR HYPERSPECTRAL IMAGE COMPRESSION

arXiv

引用

arXiv 2024年

作者： Fuchs, Martin Hermann Paul Rasti, Behnood Demir, Begüm Faculty of Electrical Engineering and Computer Science Technische Universität Berlin Germany BIFOLD - Berlin Institute for the Foundations of Learning and Data Germany

The development of learning-based hyperspectral image (HSI) compression models has recently attracted significant interest. Existing models predominantly utilize convolutional filters, which capture only local dependencies. Furthermore, they often incur high training costs and exhibit substantial computational complexity. To address these limitations, in this paper we propose Hyperspectral Compression Transformer (HyCoT) that is a transformer-based autoencoder for pixelwise HSI compression. Additionally, we apply a simple yet effective training set reduction approach to accelerate the training process. Experimental results on the HySpecNet-11k dataset demonstrate that HyCoT surpasses the state of the art across various compression ratios by over 1 dB of PSNR with significantly reduced computational requirements. Our code and pre-trained weights are publicly available at https://***/rsim/hycot. © 2024, CC BY.

关键词： Image compression

来源：评论

学校读者我要写书评

暂无评论

Neural approximation of Wasserstein distance via a universal architecture for symmetric and factorwise group invariant functions 23

Neural approximation of Wasserstein distance via a universal...

引用

Proceedings of the 37th International Conference on Neural Information Processing Systems

作者： Samantha Chen Yusu Wang Department of Computer Science Engineering University of California - San Diego Halıcıoglu Data Science Institute University of California - San Diego

Learning distance functions between complex objects, such as the Wasserstein distance to compare point sets, is a common goal in machine learning applications. However, functions on such complex objects (e.g., point sets and graphs) are often required to be invariant to a wide variety of group actions e.g. permutation or rigid transformation. Therefore, continuous and symmetric product functions (such as distance functions) on such complex objects must also be invariant to the product of such group actions. We call these functions symmetric and factor-wise group invariant functions (or SFGI functions in short). In this paper, we first present a general neural network architecture for approximating SFGI functions. The main contribution of this paper combines this general neural network with a sketching idea to develop a specific and efficient neural network which can approximate the p-th Wasserstein distance between point sets. Very importantly, the required model complexity is independent of the sizes of input point sets. On the theoretical front, to the best of our knowledge, this is the first result showing that there exists a neural network with the capacity to approximate Wasserstein distance with bounded model complexity. Our work provides an interesting integration of sketching ideas for geometric problems with universal approximation of symmetric functions. On the empirical front, we present a range of results showing that our newly proposed neural network architecture performs comparatively or better than other models (including a SOTA Siamese Autoencoder based approach). In particular, our neural network generalizes significantly better and trains much faster than the SOTA Siamese AE. Finally, this line of investigation could be useful in exploring effective neural network design for solving a broad range of geometric optimization problems (e.g., k-means in a metric space).

关键词：

来源：评论

学校读者我要写书评

暂无评论

AI-Driven Assessment and Evaluation: Enhancing Sustainability in Educational Practices

AI-Driven Assessment and Evaluation: Enhancing Sustainabilit...

引用

Automation, Computing and Renewable Systems (ICACRS), International Conference on

作者： P. Chinnasamy M. Pushpalatha V. Nirmala Vidhya V Ramesh Kumar Ayyasamy Divya Priya Department of Computer Science and Engineering School of Computing Kalasalingam Academy of Research and Education Tamilnadu India Department of Computer Science and Technology Karpagam College of Engineering Coimbatore India Department of MCA Annamacharya PG College of Computer Studies Kadapa Artificial Intelligence and Data Science Kongunadu College of Engineering and Technology (Autonomous) Faculty of Information and Communication Technology Universiti Tunku Abdul Rahman (UTAR) Kampar Perak Malaysia Department of Computer Science and Engineering MLR Institute of Technology Hyderabad India

ISBN: (数字)9798331532420

ISBN: (纸本)9798331532437

AI-powered educational technologies are emerging as transformative forces in the quickly changing field of education, where innovation is essential to keeping ahead of the competition. Assessments are one area that is undergoing revolutionary transformation. Artificial Intelligence (AI) is redefining the way we evaluate learning outcomes as traditional evaluation approaches struggle to keep up with the demands of the digital age. The field of artificial intelligence-powered learning assessments has a lot of promising future developments ahead of it. This research proposes novel technique in assessment and evaluation based on online education system using Artificial intelligence in machine learning techniques. Here, the input has been analysed as student performance data based on online education and processed for removing of missing values with noise removal. Then this data has been analysed using spatio LSTM convolutional fuzzy neural network for classification of evaluation and assessment model. the experimental analysis has been carried out for various student performance analysis dataset in terms of training accuracy, average precision, specificity, F-measure and recall. Proposed method training accuracy 97%, Average precision 94%, F-measure 90%, RECALL 93%, SPECIFICITY 96%.

关键词： Training Technological innovation Renewable energy sources Accuracy Education Learning (artificial intelligence) Machine learning Fuzzy neural networks Sustainable development Long short term memory

来源：评论

学校读者我要写书评

暂无评论

Ethereum Phishing Scams Detection Based on Graph Contrastive Learning with Augmentations

Ethereum Phishing Scams Detection Based on Graph Contrastive...

引用

International Conference on computer Supported Cooperative Work in Design

作者： Yongxin Chen Wenhan Hou Xin Zhang Ru Li School of Computer Science Inner Mongolia University Hohhot China Engineering Research Center of Ecological Big Data Ministry of Education China

ISBN: (数字)9798350349184

ISBN: (纸本)9798350349191

Cryptocurrency crime incidents in Ethereum are continuously rising, with phishing scams accounting for 50% of all criminal activities. The severe data imbalance significantly impacts the performance of Ethereum phishing detection models. The current solution may introduce redundant information or lead to the loss of important data. In this paper, we propose an Ethereum phishing detection method based on Graph Contrastive Learning with augmentations. This approach addresses the issue of insufficient learning of phishing node features, thus alleviating the influence of data imbalance on the model’s detection performance without disrupting the original data distribution. To enhance the representation of structural features, we employ two data augmentation methods: feature masking and edge perturbation. We conducted extensive experiments on a real Ethereum phishing dataset to evaluate the performance of our method. Compared to alternative methods, our approach not only significantly improves Precision, ranging from 12% to 30%, but also achieves noticeable enhancements in Recall, Auc, and F1-score. The experimental results provide ample evidence of the effectiveness of the proposed method.

关键词： Phishing Perturbation methods Image edge detection Smart contracts Contrastive learning data augmentation data models

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：