检索结果-内蒙古大学图书馆

arXiv 2022年

作者： Chai, Zhizhong Lin, Huangjing Luo, Luyang Heng, Pheng-Ann Chen, Hao Imsight AI Research Lab Shenzhen China Department of Computer Science and Engineering The Chinese University of Hong Kong Hong Kong Hong Kong Guangdong Provincial Key Laboratory of Computer Vision and Virtual Reality Technology Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences Shenzhen China Department of Computer Science and Engineering The Hong Kong University of Science and Technology Hong Kong Hong Kong

Most of the existing object detection works are based on the bounding box annotation: each object has a precise annotated box. However, for rib fractures, the bounding box annotation is very labor-intensive and time-consuming because radiologists need to investigate and annotate the rib fractures on a slice-by-slice basis. Although a few studies have proposed weakly-supervised methods or semi-supervised methods, they could not handle different forms of supervision simultaneously. In this paper, we proposed a novel omni-supervised object detection network, which can exploit multiple different forms of annotated data to further improve the detection performance. Specifically, the proposed network contains an omni-supervised detection head, in which each form of annotation data corresponds to a unique classification branch. Furthermore, we proposed a dynamic label assignment strategy for different annotated forms of data to facilitate better learning for each branch. Moreover, we also design a confidence-aware classification loss to emphasize the samples with high confidence and further improve the model’s performance. Extensive experiments conducted on the testing dataset show our proposed method outperforms other state-of-the-art approaches consistently, demonstrating the efficacy of deep omni-supervised learning on improving rib fracture detection performance. Copyright © 2022, The Authors. All rights reserved.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Modified U-Net Architeture based Ischemic Stroke Lesions Segmentation

Modified U-Net Architeture based Ischemic Stroke Lesions Seg...

引用

International Conference on Sciences and Techniques of Automatic Control and computer engineering (STA)

作者： Nabil Abdmouleh Amira Echtioui Fathi Kallel Ahmed Ben Hamida ATMS Lab Advanced Technologies for Medicine and Signals ENISSfax University Sfax Tunisia National Engineering School of Electronics and Communications Sfax University Tunisia Department IS College of Computer Science King Khaled University ‘KKU’ KSA

ISBN: (纸本)9781665482622

An ischemic stroke transpires when part of the brain tissue is unable to get oxygen and nutrients due to the interruption or the reduction of blood supply. A stroke is a medical emergency that can lead to serious complications if left untreated. Early action can reduce potential brain damage and other consequences. Therefore, quantitative assessment of ischemic stroke lesions using magnetic resonance imaging (MRI) is paramount for an accurate and decisive diagnosis. However, a diagnosis of hundreds of MRI-generated slices by trained specialists is time-consuming and requires painstaking attention to details. Automating this process using advanced image segmentation methods allows an exact localization of the tissue presenting lesions and provide an approximation of its volume. Automatic detection of lesions provides faster answers and better assessments to assist medical practitioners. In this paper, we propose a novel method for ischemic stroke lesions detection based on a modified UNet architecture. We use the data from 2015 Ischemic Stroke Lesion Segmentation Challenge (ISLES 2015) for performance evaluation of our proposed method. Experiments have shown that the latter outperforms the state-of-the-art methods.

关键词： Performance evaluation Location awareness Image segmentation Costs Magnetic resonance imaging Architecture computer architecture

来源：评论

学校读者我要写书评

暂无评论

Uncertainty Quantification of Signal Integrity Analysis for Neuromorphic Chips

IEEE Transactions on Signal and Power Integrity

引用

IEEE Transactions on Signal and Power Integrity 2022年 1卷 160-169页

作者： Hanzhi Ma Da Li Tuomin Tao Xingjian Shangguan En-Xiao Liu Jose Schutt-Aine Andreas C. Cangellaris Er-Ping Li College of Information Science and Electronic Engineering ZJU-UIUC Institute Zhejiang University Hangzhou China Key Lab of Advanced Micro/Nano Electronic Devices and Smart Systems of Zhejiang University Hangzhou China Zhejiang-Singapore Innovation and AI Joint Research Lab China Department of Electrical and Computer Engineering University of Illinois at Urbana-Champaign Urbana IL USA A*STAR Institute of High Performance Computing Singapore

A dimensionality reduction based neural network framework is introduced for uncertainty quantification of time-domain response based on system uncertain design parameters for neuromorphic chips. The proposed method firstly makes use of the singular value decomposition (SVD) method to find the basis functions and corresponding coefficients of time-domain response, of which coefficients are used as a lower dimensional target outputs in neural network model compared with time sampling points prediction. This newly proposed method then develops an integrated neural network structure to simultaneously find the mean and variance of target coefficients with a combined definition of loss function, which can be utilized together with basis functions to construct the prediction interval of time-domain response. A memrisor-based crossbar array is applied in this work to verify the performance of the proposed method with the comparison of Monte Carlo method.

关键词： Biological neural networks Uncertainty Memristors Time-domain analysis Integrated circuit modeling Neuromorphics Signal integrity Neural networks Neuromorphics

来源：评论

学校读者我要写书评

暂无评论

Observation of charge to spin conversion in p-orbital InBi alloy

arXiv

引用

arXiv 2025年

作者： Li, Gen Zhang, Ying Xu, Xiaoguang Shen, Lei Feng, Zheng Meng, Kangkang Li, Ang Cheng, Lu He, Kang Tan, Wei Wu, Yong Wu, Yihong Jiang, Yong Key Laboratory of Advanced Materials and Devices for Post-Moore Chips Ministry of Education School of Materials Science and Engineering University of Science and Technology Beijing Beijing100083 China Department of Electrical and Computer Engineering National University of Singapore 117576 Singapore Research Institute Chongqing Liang Jiang New Area Chongqing401123 China Department of Mechanical Engineering National University of Singapore 117575 Singapore Microsystem & Terahertz Research Center CAEP Chengdu610200 China Faculty of Materials and Manufacturing Beijing Key Lab of Microstructure and Properties of Advanced Materials Beijing University of Technology Beijing100124 China Institute of Quantum Materials and Devices School of Electronic and Information Engineering State Key Laboratory of Separation Membranes and Membrane Processes Tiangong University Tianjin300387 China

High density data storage and spin-logic devices require highly efficient all-electric control of spin moment. So far, charge-to-spin conversion through the spin Hall effect limits to d-orbital materials associated with strong spin-orbit coupling, especially heavy metals. However, d-orbital heavy metals with strong spin-orbit coupling results in a short spin diffusion length, which restricts the spin transport and accumulation in spintronic devices. Therefore, the conflict between charge-to-spin conversion efficiency and spin transport ability is irreconcilable in d-orbital materials. Here, we report a large charge to spin conversion experimentally observed in the p-orbital In2Bi alloy, exhibiting the coexistence of a large spin Hall angle comparable to heave metal Pt and a long spin diffusion length (4 times that of Pt). First-principles calculations reveal that topological symmetry-protected gap opening near the Fermi level results in large Berry curvature-related spin Hall conductivity. Due to the delocalized nature of p-orbitals and semimetal properties of In2Bi, its spin current can overcome the physical barriers between spin Hall angle and spin diffusion length in d-orbital metals, thereby advancing the development of high-performance spintronic devices. Copyright © 2025, The Authors. All rights reserved.

关键词： Heavy metals

来源：评论

学校读者我要写书评

暂无评论

FCloudless: A Performance-Aware Collaborative Mechanism for JointCloud Serverless

FCloudless: A Performance-Aware Collaborative Mechanism for ...

引用

IEEE International Conference on Joint Cloud Computing (JCC)

作者： Jianfei Liu Huaimin Wang Peichang Shi Yaojie Li Penghui Ma Guodong Yi National Key Laboratory of Parallel and Distributed Computing College of Computer Science National University of Defense Technology Changsha 410073 China Key Laboratory of Software Engineering for Complex Systems College of Computer Science National University of Defense Technology Changsha 410073 China Xiangjiang Lab Changsha 410073 China School of Advanced Interdisciplinary Studies Hunan University Of Technology and Business Changsha 410073 China

As a new stage in the development of the cloud computing paradigm, serverless computing has the high-level abstraction characteristic of shielding underlying details. This makes it extremely challenging for users to choose a suitable serverless platform. To address this, targeting the jointcloud computing scenario of heterogeneous serverless platforms across multiple clouds, this paper presents a jointcloud collaborative mechanism called FCloudless with cross-cloud detection of the full lifecycle performance of serverless platforms. Based on the benchmark metrics set that probe performance critical stages of the full lifecycle, this paper proposes a performance optimization algorithm based on detected performance data that takes into account all key stages that affect the performance during the lifecycle of a function and predicts the overall performance by combining the scores of local stages and dynamic weights. We evaluate FCloudless on AWS, AliYun, and Azure. The experimental results show that FCloudless can detect the underlying performance of serverless platforms hidden in the black box and its optimization algorithm can select the optimal scheduling strategy for various applications in a jointcloud environment. FCloudless reduces the runtime by 23.3% and 24.7% for cold and warm invocations respectively under cost constraints.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Hyperbolic geometric latent diffusion model for graph generation 24

Hyperbolic geometric latent diffusion model for graph genera...

引用

Proceedings of the 41st International Conference on Machine Learning

作者： Xingcheng Fu Yisen Gao Yuecen Wei Qingyun Sun Hao Peng Jianxin Li Xianxian Li Key Lab of Education Blockchain and Intelligent Technology Ministry of Education Guangxi Normal University Guilin China Institute of Artificial Intelligence Beihang University Beijing China and Key Lab of Education Blockchain and Intelligent Technology Ministry of Education Guangxi Normal University Guilin China School of Software Beihang University Beijing China Beijing Advanced Innovation Center for Big Data and Brain Computing School of Computer Science and Engineering Beihang University Beijing China

Diffusion models have made significant contributions to computer vision, sparking a growing interest in the community recently regarding the application of them to graph generation. Existing discrete graph diffusion models exhibit heightened computational complexity and diminished training efficiency. A preferable and natural way is to directly diffuse the graph within the latent space. However, due to the non-Euclidean structure of graphs is not isotropic in the latent space, the existing latent diffusion models effectively make it difficult to capture and preserve the topological information of graphs. To address the above challenges, we propose a novel geometrically latent diffusion framework HypDiff. Specifically, we first establish a geometrically latent space with interpretability measures based on hyperbolic geometry, to define anisotropic latent diffusion processes for graphs. Then, we propose a geometrically latent diffusion process that is constrained by both radial and angular geometric properties, thereby ensuring the preservation of the original topological properties in the generative graphs. Extensive experimental results demonstrate the superior effectiveness of HypDiff for graph generation with various topologies.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Free form three dimensional integrated circuits and wearables on a thread using organic eutectogel gated electrochemical transistors

arXiv

引用

arXiv 2023年

作者： Owyeung, Rachel E. Zeng, Wenxin Panzer, Matthew J. Sonkusale, Sameer Department of Chemical and Biological Engineering Tufts University 4 Colby Street MedfordMA02155 United States Department of Electrical and Computer Engineering Tufts University 161 College Ave MedfordMA02155 United States Nano Lab Advanced Technology Laboratory Tufts University 200 Boston Ave. Suite 2600 MedfordMA02155 United States

Existing flexible electronics are planar, nonbreathable, and easily damaged. In this work, authors introduce a new paradigm of free-form three-dimensional integrated circuits assembled on a single textile thread for ultimate flexibility. Transistors are assembled as daisy chain on a thread and utilize deep eutectic solvent gels, or eutectogels as nonvolatile, breathable, and repairable dielectric for gating organic electrochemical transistor. The eutectogels provide stable transistor performance without the need for encapsulation, are air permeable to the skin, and make the Organic Eutectogel Gated Electrochemical Transistor (OEGET) a promising candidate for wearable electronics. Complex analog integrated circuits, such as single stage and multi-stage amplifiers can be realized on a single thread which can bend out of the plane of stretching for true three-dimensional flexibility. An all-thread based wearable that uses a thread-based strain sensor and thread-based integrated circuit for eye motion and respiration monitoring is also presented. © 2023, CC BY-NC-ND.

关键词： Flexible electronics

来源：评论

学校读者我要写书评

暂无评论

Dynamic resolution network 21

Dynamic resolution network

引用

Proceedings of the 35th International Conference on Neural Information Processing Systems

作者： Mingjian Zhu Kai Han Enhua Wu Qiulin Zhang Ying Nie Zhenzhong Lan Yunhe Wang Zhejiang University and Huawei Noah's Ark Lab. and School of Engineering Westlake University and Institute of Advanced Technology Westlake Institute for Advanced Study Huawei Noah's Ark Lab. and State Key Lab of Computer Science ISCAS & University of Chinese Academy of Sciences State Key Lab of Computer Science ISCAS & University of Chinese Academy of Sciences and University of Macau BUPT Huawei Noah's Ark Lab. School of Engineering Westlake University and Institute of Advanced Technology Westlake Institute for Advanced Study

ISBN: (纸本)9781713845393

Deep convolutional neural networks (CNNs) are often of sophisticated design with numerous learnable parameters for the accuracy reason. To alleviate the expensive costs of deploying them on mobile devices, recent works have made huge efforts for excavating redundancy in pre-defined architectures. Nevertheless, the redundancy on the input resolution of modern CNNs has not been fully investigated, i.e., the resolution of input image is fixed. In this paper, we observe that the smallest resolution for accurately predicting the given image is different using the same neural network. To this end, we propose a novel dynamic-resolution network (DRNet) in which the input resolution is determined dynamically based on each input sample. Wherein, a resolution predictor with negligible computational costs is explored and optimized jointly with the desired network. Specifically, the predictor learns the smallest resolution that can retain and even exceed the original recognition accuracy for each image. During the inference, each input image will be resized to its predicted resolution for minimizing the overall computation burden. We then conduct extensive experiments on several benchmark networks and datasets. The results show that our DRNet can be embedded in any off-the-shelf network architecture to obtain a considerable reduction in computational complexity. For instance, DR-ResNet-50 achieves similar performance with an about 34% computation reduction, while gaining 1.4% accuracy increase with 10% computation reduction compared to the original ResNet-50 on ImageNet.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Vote and KNN outlier detection in Wireless Sensor Networks

Vote and KNN outlier detection in Wireless Sensor Networks

引用

International Symposium on Networks, computers and Communications (ISNCC)

作者： Aymen Abid Salim El Khediri Tarek Moulahi Rym Cheour Abdennaceur Kachouri Computer and Embedded Systems Lab National Engineering School Sfax TUNISIA College of Computer Qassim University Qassim KSA Laboratoire HATEM Bettaher (IRESCOMATH) Gabes Tunisia Advanced Fluid Dynamics Energetics and Evironment Lab National Engineering School Sfax TUNISIA

The target of a Wireless Sensor Network (WSN) designer is to improve the robustness of the network while taking into account constraints such as verifying the detection of anomalies. Preventing anomalies can be assured by data analysis via outliers identification as it strongly affects decision-making in the case of smart-home, e-health and other high-risk applications. In this paper, we study some families of outliers and we propose to detect data anomalies using K Nearest Neighbours (KNN) and Voting techniques by a benchmark suitable to be a simulation for a smart-house. In order to evaluate the dependability, some performance metrics are analyzed such as data rate detection, precision, specificity and accuracy. Performance results are in the order of 100% detection rate with reduced false alarm and fast response time compared to some other existing ones.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Learning contrastive embedding in low-dimensional space 22

Learning contrastive embedding in low-dimensional space

引用

Proceedings of the 36th International Conference on Neural Information Processing Systems

作者： Shuo Chen Chen Gong Jun Li Jian Yang Gang Niu Masashi Sugiyama RIKEN Center for Advanced Intelligence Project (AIP) Japan PCA Lab Key Lab of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education and Jiangsu Key Lab of Image and Video Understanding for Social Security School of Computer Science and Engineering Nanjing University of Science and Technology China RIKEN Center for Advanced Intelligence Project (AIP) Japan and Graduate School of Frontier Sciences The University of Tokyo Japan

ISBN: (纸本)9781713871088

Contrastive learning (CL) pretrains feature embeddings to scatter instances in the feature space so that the training data can be well discriminated. Most existing CL techniques usually encourage learning such feature embeddings in the high-dimensional space to maximize the instance discrimination. However, this practice may lead to undesired results where the scattering instances are sparsely distributed in the high-dimensional feature space, making it difficult to capture the underlying similarity between pairwise instances. To this end, we propose a novel framework called contrastive learning with low-dimensional reconstruction (CLLR), which adopts a regularized projection layer to reduce the dimensionality of the feature embedding. In CLLR, we build the sparse/low-rank regularizer to adaptively reconstruct a low-dimensional projection space while preserving the basic objective for instance discrimination, and thus successfully learning contrastive embeddings that alleviate the above issue. Theoretically, we prove a tighter error bound for CLLR; empirically, the superiority of CLLR is demonstrated across multiple domains. Both theoretical and experimental results emphasize the significance of learning low-dimensional contrastive embeddings.

关键词：

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：