检索结果-内蒙古大学图书馆

IEEE International Conference on Communications in China (ICCC)

作者： Minghong Wu Minghui Liwang Yuhan Su Yuliang Tang Zhenzhen Jiao Xianbin Wang School of Informatics Xiamen University Xiamen China College of Electronic and Information Engineering Tongji University Shanghai China School of Electronic Science and Engineering Xiamen University Xiamen China iF-Labs Beijing Teleinfo Technology Co. Ltd. CAICT China Dept. of Electrical and Computer Engineering Western University Ontario Canada

ISBN: (数字)9798350378412

ISBN: (纸本)9798350378429

Federated Learning (FL) represents a popular distributed learning architecture that facilitates data privacy by enabling clients (e.g., mobile devices) to train a FL model collaboratively without data sharing. Existing efforts mainly focus on accuracy, delay and energy consumption over a rather stable network with certain clients, with less emphasis on the impact of frequent decision-making processes during model training. Inspired by this, this paper considers a dynamic FL network with uncertain participation of clients, while we jointly optimize client selection and communication resource allocation to achieve a balance between energy cost and FL model accuracy, as proved to be NP-hard. To facilitate timely model training, we propose a prediction-based two-stage asynchronous programming mechanism, which decouples the problem in two subproblems, corresponding to two stages. In particular, the former stage determines some long-term clients which are more stable to join in prior to practical model training process, by estimating the online probability of clients. Then, the latter stage can be implemented by involving some temporary clients as backups when long-term ones are not able to show up. Such a well-designed mechanism offers a unique veiw on FL, while enabling a responsive and cost-effective decision-making process. Comprehensive simulations regarding both IID and non-IID data distributions on MNIST and CIFAR-10 datasets can prove our commendable performance on time efficiency, energy cost and accuracy.

关键词： Training Accuracy Costs Federated learning Decision making Programming Dynamic scheduling

来源：评论

学校读者我要写书评

暂无评论

Neural Annotation Refinement: Development of a New 3D Dataset for Adrenal Gland Analysis

arXiv

引用

arXiv 2022年

作者： Yang, Jiancheng Shi, Rui Wickramasinghe, Udaranga Zhu, Qikui Ni, Bingbing Fua, Pascal Shanghai Jiao Tong University Shanghai China EPFL Lausanne Switzerland Dept. of Computer and Data Science Case Western Reserve University OH United States Dept. of Biomedical Engineering Case Western Reserve University OH United States

The human annotations are imperfect, especially when produced by junior practitioners. Multi-expert consensus is usually regarded as golden standard, while this annotation protocol is too expensive to implement in many real-world projects. In this study, we propose a method to refine human annotation, named Neural Annotation Refinement (NeAR). It is based on a learnable implicit function, which decodes a latent vector into represented shape. By integrating the appearance as an input of implicit functions, the appearance-aware NeAR fixes the annotation artefacts. Our method is demonstrated on the application of adrenal gland analysis. We first show that the NeAR can repair distorted golden standards on a public adrenal gland segmentation dataset. Besides, we develop a new Adrenal gLand ANalysis (ALAN) dataset with the proposed NeAR, where each case consists of a 3D shape of adrenal gland and its diagnosis label (normal vs. abnormal) assigned by experts. We show that models trained on the shapes repaired by the NeAR can diagnose adrenal glands better than the original ones. The ALAN dataset will be open-source, with 1,584 shapes for adrenal gland diagnosis, which serves as a new benchmark for medical shape analysis. Code and dataset are available at https://***/M3DV/NeAR. © 2022, CC BY.

关键词： Diagnosis

来源：评论

学校读者我要写书评

暂无评论

Evaluation of the Xilinx Deep Learning Processing Unit under Neutron Irradiation

Evaluation of the Xilinx Deep Learning Processing Unit under...

引用

European Conference on Radiation and its Effects on Components and Systems (RADECS)

作者： Dimitris Agiakatsikas Nikos Foutris Aitzan Sari Vasileios Vlagkoulis Ioanna Souvatzoglou Mihalis Psarakis Mikel Luján Maria Kastriotou Carlo Cazzaniga Dept. of Informatics University of Piraeus Greece Dept. of Computer Science The University of Manchester UK Isis Facility Stfc Rutherford Appleton Laboratory UK

This paper studies the dependability of the Xilinx Deep-Learning Processing Unit (DPU) under neutron irradiation. It analyses the impact of Single Event Effects (SEEs) on the accuracy of the DPU running the resnet50 m... 详细信息

ISBN: (数字)9781665437943

ISBN: (纸本)9781665437950

关键词： Deep learning Radiation effects Full stack Failure analysis Europe Neutrons Reliability

来源：评论

学校读者我要写书评

暂无评论

Explainable Feature Embeddings from Histopathology Foundation Models: A Case Study for End Stage Kidney Disease Risk Analysis in Diabetic Nephropathy Patients

Explainable Feature Embeddings from Histopathology Foundatio...

引用

Medical Imaging 2025: Digital and Computational Pathology

作者： Kasireddy, Harishwar Reddy Lucarelli, Nicholas Yun, Donghwan Moon, Kyung Chul La Rosa, Patricio S. Tomaszewski, John E. Han, Seung Seok Shickel, Benjamin Naglah, Ahmed Sarder, Pinaki Dept. of Electrical and Computer Engineering Univ. of Florida GainesvilleFL United States Univ. of Florida Intelligent Critical Care Center GainesvilleFL United States Dept. of Biomedical Engineering Univ. of Florida GainesvilleFL United States Dept. of Internal Medicine Seoul National Univ. Seoul Korea Republic of Dept. of Pathology Seoul National Univ. Seoul Korea Republic of Seed Production Innovation Crop Science Division Bayer Company St. LouisMO United States Dept. of Pathology and Anatomical Sciences Univ. at Buffalo BuffaloNY United States Division of Nephrology Hypertension and Renal Transplantation Quantitative Health Section Dept. of Medicine Univ. of Florida GainesvilleFL United States

ISBN: (纸本)9781510686045

Foundational models (FMs) based on advanced neural network architectures have demonstrated improved performance in pathology image analysis across various organs due to their increased generalizability. However, their clinical adoption requires explainability, as their black box nature limits transparency. Understanding the specific features these models learn for a given downstream task is crucial for explainability and integrating FMs into clinical workflows more effectively. We propose a computational pipeline that enhances explainability by correlating domain-specific handcrafted features (HFs), with hidden features i.e., feature embeddings (FEs) from FMs. We correlate and combine HFs from Detectron 2 DeepLabv3+ segmentation with FEs from Prov-Gigapath (PG) and UNI FMs for improved explainability and performance. In this work, HFs are extracted from segmented functional tissue units, including arteries, tubules, globally sclerotic glomeruli, and non-globally sclerotic glomeruli. FEs are extracted at the tile and slide levels for PG and at the tile level for UNI. We use the Pearson correlation coefficient to identify significant correspondences between these feature sets. To evaluate our proposed methodology, we use 56 diabetic nephropathy kidney biopsy whole slide images (WSIs) from Seoul National University Hospital. The task is to predict end-stage kidney disease (ESKD) two years post-biopsy using leave-one-out cross-validation on 56 WSIs, with 16 from ESKD patients and 40 from non-ESKD patients. We combine top correlated features from FEs of FMs with HFs and train logistic regression (LR) and k nearest neighbor (kNN) classifiers. LR model trained on combined feature set improved accuracy, balanced accuracy, Matthew's correlation coefficient, F1-score, precision, and recall to 0.8393, 0.7938, 0.5993, 0.8377, 0.8367, 0.8393 respectively, when compared to LR and kNN models trained on individual feature sets. PG excelled in specificity (1.000) and AUROC (0.8281),

关键词： Biopsy

来源：评论

学校读者我要写书评

暂无评论

Effect of Interest Variation and Prefix Length on Pending Interest Table (PIT) Performance in Named Data Networking (NDN)

Effect of Interest Variation and Prefix Length on Pending In...

引用

International Conference on Wireless and Telematics (ICWT)

作者： Helmi Ahmad Khaderani Ishlah Muhammadi Lutfi Razan Muhammad Adi Sucipto Luthfi Restu Kusumastuti Ade Nurhayati Syaiful Ahdan Nana Rachmana Syambas Dept. of Telecommunication Engineering Bandung Institute of Technology Bandung Indonesia School of Electrical Engineering and Informatics Bandung Institute of Technology Bandung Indonesia Faculty of Engineering & Computer Science Universitas Teknokrat Indonesia Lampung Indonesia Faculty of Engineering Telkom University Jakarta Barat Indonesia

Pending interest table (PIT) is one of the data structures on each named data network router. The speed of delivery of interest as well as the interest served at PIT are several important parameters for measuring PIT performance. This research was conducted to analyze the performance of PIT by measuring the round trip time (RTT) of sending data and sending interests from consumers to producers on the Palapa Ring Sumatra topology. Testing is carried out using miniNDN by varying the length of the prefix and the number of interests sent. From the simulation and results of the analysis that has been carried out, it is known that the correlation between the length of the prefix and RTT is directly proportional. The longer the prefix of an interest packet, the higher the RTT value. Apart from the simulation results, the number of interest packets sent is the same as those received at each node.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Video Streaming QoS Analysis with TIPHON Standard NDNts

Video Streaming QoS Analysis with TIPHON Standard NDNts

引用

International Conference on Wireless and Telematics (ICWT)

作者： I Made Aditya Putra J.D Lisdatul Alifa Muhammad Hilmi Fawwaz Nissha Kinaya Mauren Sibayak Ratna Mayasari Hasan Nur Arifin Syaiful Ahdan Ridha Muldina Negara Nana Rachmana Syambas Dept. of Telecommunication Engineering Bandung Institute of Technology Bandung Indonesia School of Electrical Engineering and Informatics Bandung Institute of Technology Bandung Indonesia School of Electrical Engineering Telkom University Bandung Indonesia Faculty of Engineering & Computer Science Universitas Teknokrat Indonesia Lampung Indonesia

Named data networks are a new concept in network architecture that can process data transmission by name. This concept can be utilized in the case of video streaming. TIPHON is a standard used to support streaming multimedia activities. According to TIPHON, three factors-delay, packet loss, and throughput-can influence how a stream behaves. This study's goal is to evaluate the quality of service (QoS) of video streaming services using an NDN network that adheres to the TIPHON NDN standard. Several things can affect throughput, such as bottlenecks on the computer or network device used.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Bayesian modeling of effective and functional brain connectivity using hierarchical vector autoregressions

arXiv

引用

arXiv 2021年

作者： Wegmann, Bertil Lundquist, Anders Eklund, Anders Villani, Mattias Div. of Statistics and Machine Learning Dept. of Computer and Information Science Linköping University Sweden Umeå University Sweden Div. of Medical Informatics Dept. of Biomedical Engineering Linköping University Sweden Dept. of Statistics Stockholm University Sweden

Analysis of brain connectivity is important for understanding how information is processed by the brain. We propose a novel Bayesian vector autoregression (VAR) hierarchical model for analyzing brain connectivity in a resting-state fMRI data set with autism spectrum disorder (ASD) patients and healthy controls. Our approach models functional and effective connectivity simultaneously, which is new in the VAR literature for brain connectivity, and allows for both group- and single-subject inference as well as group comparisons. We combine analytical marginalization with Hamiltonian Monte Carlo (HMC) to obtain highly efficient posterior sampling. The results from more simplified covariance settings are, in general, overly optimistic about functional connectivity between regions compared to our results. In addition, our modeling of heterogeneous subject-specific covariance matrices is shown to give smaller differences in effective connectivity compared to models with a common covariance matrix to all subjects. © 2021, CC BY.

关键词： Covariance matrix

来源：评论

学校读者我要写书评

暂无评论

A new clustering-based technique for the acceleration of deep convolutional networks

arXiv

引用

arXiv 2021年

作者： Pikoulis, Erion Vasilis Mavrokefalidis, Christos Lalos, Aris S. Industrial Systems Institute Patras Science Park Greece Computer Engineering Informatics Dept. University of Patras Greece

Deep learning and especially the use of Deep Neural Networks (DNNs) provides impressive results in various regression and classification tasks. However, to achieve these results, there is a high demand for computing and storing resources. This becomes problematic when, for instance, real-time, mobile applications are considered, in which the involved (embedded) devices have limited resources. A common way of addressing this problem is to transform the original large pre-trained networks into new smaller models, by utilizing Model Compression and Acceleration (MCA) techniques. Within the MCA framework, we propose a clustering-based approach that is able to increase the number of employed centroids/representatives, while at the same time, have an acceleration gain compared to conventional, k-means based approaches. This is achieved by imposing a special structure to the employed representatives, which is enabled by the particularities of the problem at hand. Moreover, the theoretical acceleration gains are presented and the key system hyper-parameters that affect that gain, are identified. Extensive evaluation studies carried out using various state-of-the-art DNN models trained in image classification, validate the superiority of the proposed method as compared for its use in MCA tasks. © 2021, CC BY.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

People movement linkage based on path revision across multiple cameras

People movement linkage based on path revision across multip...

引用

作者： Chang, I-Cheng Liu, Chieh-Yu Huang, Chung-Lin Kabi, Kunal Dept. of Computer Science and Information Engineering National Dong Hwa University Taiwan Dept. of Electrical Engineering National Tsing Hua University Taiwan Dept. of Applied Informatics and Multimedia Asia University Taiwan

This work presents an object-based people motion linkage system, which tracks and records the behavior of each person across multiple cameras. The proposed system includes two phases: path construction phase and path revision phase. The spatiotemporal relationships among cameras are trained by batch-learning procedures, and the appearance model is improved by color calibration for different cameras. When a person moves across cameras, the spatiotemporal relationship and appearance model are used to obtain the correspondence. However, object tracking may be lost due to some unexpected events. We revise the tracking paths by using the backward tracking technique to connect the missed links. Moreover, the trained Hidden Markov Models are further used to reconstruct the normal paths of these objects. In the experimental results, we demonstrate the efficiency of our approach by showing the reconnection of missing paths under different conditions. © 2020 Taiwan Academic Network Management Committee. All rights reserved.

关键词： Dynamic programming

来源：评论

学校读者我要写书评

暂无评论

Editorial

Recent Advances in Computer Science and Communications

引用

Recent Advances in computer science and Communications 2021年第5期14卷 1476-1476页

作者： Thirumoorthy, Gopalakrishnan Priya, R. Vidhya School of Computer Science and Engineering Vellore Institute of Technology Vellore India Dept. of Biomedical Engineering PSG College of Engineering Coimbatore India

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：