检索结果-内蒙古大学图书馆

LayerCFL:an efcient federated learning with layer-wised clustering

Cybersecurity 2025年第1期8卷 72-85页

作者： Jie Yuan Rui Qian Tingting Yuan Mingliang Sun Jirui Li Xiaoyong Li School of Cyberspace Security Beijing University of Posts and TelecommunicationsBeijingChina Key Laboratory of Trustworthy Distributed Computing and Service(BUPT) Ministry of EducationBeijingChina Institute of Computer Science Faculty of Mathematics and Computer ScienceUniversity of GoettingenGöettingenGermany School of Information Technology Henan University of Chinese MedicineZhengzhouHenanChina

Federated Learning(FL)sufers from the Non-IID problem in practice,which poses a challenge for efcient and accurate model *** address this challenge,prior research has introduced clustered FL(CFL),which involves clustering clients and training them *** its potential benefts,CFL can be computationally and communicationally expensive when the data distribution is unknown *** is because CFL involves the entire neural networks of involved clients in computing the clusters during training,which can become increasingly timeconsuming with large-sized *** tackle this issue,this paper proposes an efcient CFL approach called LayerCFL that employs a Layer-wised clustering *** LayerCFL,clients are clustered based on a limited number of layers of neural networks that are pre-selected using statistical and experimental *** experimental results demonstrate the efectiveness of LayerCFL in mitigating the impact of Non-IID data,improving the accuracy of clustering,and enhancing computational efciency.

关键词： Federated learning Clustered federated learning Non-IID Layer-wised clustering

来源：评论

学校读者我要写书评

暂无评论

NeuVSA: A Unified and Efficient Accelerator for Neural Vector Search 31

NeuVSA: A Unified and Efficient Accelerator for Neural Vecto...

引用

31st IEEE International Symposium on High Performance computer architecture, HPCA 2025

作者： Yuan, Ziming Dai, Lei Li, Wen Zhang, Jie Liang, Shengwen Wang, Ying Liu, Cheng Li, Huawei Li, Xiaowei Guo, Jiafeng Wang, Peng Chen, Renhai Zhang, Gong Shanxi University School of Computer and Information Technology China Peking University School of Computer Science China Institute of Computing Technology CAS State Key Lab of Processors China University of Chinese Academy of Sciences China Zhongguancun National Laboratory Beijing China Institute of Computing Technology CAS Key Lab of Network Data Science and Technology China Huawei Technologies Co. Ltd. China

ISBN: (纸本)9798331506476

Neural Vector Search (NVS) has exhibited superior search quality over traditional key-based strategies for information retrieval tasks. An effective NVS architecture requires high recall, low latency, and high throughput to enhance user experience and cost-efficiency. However, implementing NVS on existing neural network accelerators and vector search accelerators is sub-optimal due to the separation between the embedding stage and vector search stage at both algorithm and architecture levels. Fortunately, we unveil that Product Quantization (PQ) opens up an opportunity to break separation. However, existing PQ algorithms and accelerators still focus on either the embedding stage or the vector search stage, rather than both simultaneously. Simply combining existing solutions still follows the beaten track of separation and suffers from insufficient parallelization, frequent data access conflicts, and the absence of scheduling, thus failing to reach optimal recall, latency, and throughput. To this end, we propose a unified and efficient NVS accelerator dubbed NeuVSA based on algorithm and architecture co-design philosophy. Specifically, on the algorithm level, we propose a learned PQ-based unified NVS algorithm that consolidates two separate stages into the same computing and memory access paradigm. It integrates an end-to-end joint training strategy to learn the optimal codebook and index for enhanced recall and reduced PQ complexity, thus achieving smoother acceleration. On the architecture level, we customize a homogeneous NVS accelerator based on the unified NVS algorithm. Each sub-accelerator is optimized to exploit all parallelism exposed by unified NVS, incorporating a structured index assignment strategy and an elastic on-chip buffer to alleviate buffer conflicts for reduced latency. All sub-accelerators are coordinated using a hardware-aware scheduling strategy for boosted throughput. Experimental results show that the joint training strategy improves recall

关键词： Network embeddings

来源：评论

学校读者我要写书评

暂无评论

k-TEVS: A k-Times E-Voting Scheme on Blockchain With Supervision

引用

IEEE Transactions on Dependable and Secure computing 2025年第3期22卷 2326-2337页

作者： Liu, Yang He, Debiao Luo, Min Wang, Lianhai Peng, Cong Wuhan University School of Cyber Science and Engineering Wuhan430072 China Shenzhen Institute of Information Technology Institute of Information Technology Shenzhen518172 China Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center Jinan250014 China Matrix Elements Technologies Shanghai Technology Innovation Centre of Distributed Privacy-Preserving Artificial Intelligence Shanghai200232 China Shandong Fundamental Research Center for Computer Science Shandong Provincial Key Laboratory of Industrial Network and Information System Security Jinan250014 China

Thee-vote is regarded as a waytoexpress the opinion that the voters ask for. Actually, the e-vote could be applied wildly likequestionnaire,***,thecoexistences of efficiency and security as well as transparency and privacy could be considered as building blocks in the e-vote system. The blockchain could provide a public access board to reduce the storagecostsforthefieldconsistedofthevotegroupmanager(GM) with its vote assistants (VA). Particularly the k-times anonymous authentication (k-TAA) could also be a practical approach to pre serve voters’ privacy and reduce the computation costs during the ***,thee-voteschemewithpurek-TAAstrategy could damage either the supervision of voting or the efficiency and consistency of authentication process. What’s more, the impacts of dishonest voters couldn’t be stopped until the vote end. To tackle these problems, we apply the accumulator technology to add or revoke the voters at any time and extend the framework of k-TAA with the update process for the e-vote on blockchain for supervision (k-TEVS). In our scheme, the voter updates his membership witness and proves the fact that he is still a valid member with respective VA under the latest accumulator value. What’s more, this witness update operation is not contained in the authentication process, which means that the authentication process is still constant and efficient. Moreover, our add or delete update process with signature of knowledge needs only one pairing operation. For the security, we prove that the relaxed anonymity still holds in the k-TEVS framework. Finally, We implement k TEVS scheme, the Emura’s work [1] and the Huang’s work [2] for comparison. Then we make time cost and communication cost experiments, which present the feasibility and practicality of this scheme. © 2004-2012 IEEE.

关键词： Efficiency

来源：评论

学校读者我要写书评

暂无评论

Image Generation Method for Addressing Class Imbalance in Small-Sample Pulsar Candidates 21st

Image Generation Method for Addressing Class Imbalance in ...

引用

21st Pacific Rim International Conference on Artificial Intelligence, PRICAI 2024

作者： Zhang, Maoyu Xu, Hai Yan, Fanfan Ding, Haoran Guo, Meng Shandong Jinan250013 China Shandong Provincial Key Laboratory of Computing Power Internet and Service Computing Shandong Fundamental Research Center for Computer Science Shandong Jinan250013 China Jinan Institute of Supercomputing Technology Shandong Jinan250013 China

ISBN: (纸本)9789819601219

Pulsars, rapidly rotating neutron stars, have been crucial topics in astrophysics since their discovery. The advent of deep learning offers a novel method for the efficient classification of pulsar candidates. However, the extreme imbalance between positive and negative classes in the dataset severely affects the performance of the classification model. Therefore, a network enhancing detail perception and stability (EDPSNet) is proposed to address class imbalance. The model combines Deep Convolutional Generative Adversarial Networks (DCGAN) with multi-scale Channel-Spatial Attention (CSA) modules and Spectral Normalization (SN). To enhance local details and global features, thereby enhancing the overall quality and detail of generated images, we introduced a self-designed CSA module. To enhance the stability of the model, SN stabilizes training and prevents mode collapse by constraining the weight matrices of the network. In addition, we proposed the Pixel Moving Segmentation (PMS) technique to address the overfitting issue in generative models caused by insufficient samples. This method expands the dataset while preserving the original features, thus increasing the sample diversity. Experimental results indicate that the EDPSNet model and PMS technique offer significant advantages in handling class imbalanced, small-sample pulsar candidate datasets. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Pulsars

来源：评论

学校读者我要写书评

暂无评论

Expressive Speech Synthesis Enhancement with Conditional Embeddings 21st

Expressive Speech Synthesis Enhancement with Conditional Em...

引用

21st Pacific Rim International Conference on Artificial Intelligence, PRICAI 2024

作者： Yan, Fanfan Zhang, Maoyu Xu, Hai Ding, Haoran Guo, Meng Shandong Jinan250013 China Shandong Provincial Key Laboratory of Computing Power Internet and Service Computing Shandong Fundamental Research Center for Computer Science Shandong Jinan250013 China Jinan Institute of Supercomputing Technology Shandong Jinan250013 China

ISBN: (纸本)9789819601240

In recent years, emotional speech synthesis techniques have attracted considerable interest because of their wide-ranging potential applications. However, when confronted with datasets containing emotional attributes, speech synthesized by traditional methods frequently encounters difficulties, such as a mismatch with the text content or an unnatural expression of emotions. To solve these problems, we developed a straightforward emotional speech synthesis model. This model builds upon the VITS framework and incorporates an emotion prediction module, a prosody prediction module, and a conditional encoder. It can automatically predict emotion labels based on the input text, or manually specify emotions for precise control of emotions. Experimental outcomes indicate improved naturalness and expressiveness in synthesized speech, thus enhancing the overall audio quality. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Speech enhancement

来源：评论

学校读者我要写书评

暂无评论

A Weakly Supervised Semantic Segmentation Model with Enhanced CLIP Feature Extraction

A Weakly Supervised Semantic Segmentation Model with Enhance...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Kong, Fanxuan Lu, Jun College of Computer Science and Technology Heilongjiang University Harbin150080 China Jiaxiang Industrial Technology Research Institute of Heilongjiang University Jining272400 China Key Laboratory of Database and Parallel Computing of Heilongjiang Province Harbin150080 China

ISBN: (纸本)9798350368741

This paper addresses the limitations of the Contrastive Language-Image Pre-training (CLIP) model's image encoder and proposes a segmentation model WSSS-ECFE with enhanced CLIP feature extraction, aiming to improve the performance of the Weakly Supervised Semantic Segmentation (WSSS) task. WSSS-ECFE employs the Enhanced Bottleneck module proposed in this paper and adds dynamic residual connection to improve the model's processing effect on complex scenes. In terms of implementation, the Enhanced Bottleneck module employs the Swish activation function and the Depthwise Separable Convolution to enhance the feature extraction and segmentation capability of the model, and uses multiple attention mechanisms to further optimize the feature representation and segmentation accuracy. The WSSS task on the public datasets PASCAL VOC 2012 and MS COCO 2014 achieves 82.6% and 56.3% mean intersection over union (mIoU), achieving state-of-the-art performance in models with low resource requirements. © 2025 IEEE.

关键词： Attention Mechanism Bottleneck Feature Extraction Residual Connection Weakly Supervised Semantic Segmentation (WSSS)

来源：评论

学校读者我要写书评

暂无评论

Dual Space Representation Learning for Skeleton-based Action Recognition

引用

IEEE Signal Processing Letters 2025年 32卷 2104-2108页

作者： Yang, Yuheng Chen, Haipeng Liu, Zhenguang Hu, Sihao Jiao, Yingying Jilin University College of Computer Science and Technology Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Changchun130012 China Zhejiang University School of Cyber Science and Technology Hangzhou310007 China Georgia Institute of Technology College of Computing GA United States

Skeleton-based action recognition is crucial for machine intelligence. Current methods generally learn from 3D articulated motion sequences in the straightforward Euclidean space. Yet, the vanilla Euclidean space may not be the optimal choice for modeling the intricate correlations among human body joints. This challenge arises from the non-Euclidean nature of human anatomy, where joint correlations often vary non-linearly during movement. To address this, we propose a dual space representation learning method. Specifically, we represent the motion sequences in Hyperbolic space, leveraging its intrinsic properties to capture the non-Euclidean latent anatomy of human motions. We then incorporate the motion features from both Hyperbolic and Euclidean spaces, allowing us to precisely model the non-linear joint correlations while effectively sketching human poses. The proposed method empirically achieves state-of-the-art performance on the NTU RGB+D 60, NTURGB+D 120, and NW-UCLA datasets. © 1994-2012 IEEE.

关键词： Digital arithmetic

来源：评论

学校读者我要写书评

暂无评论

CurMIM: Curriculum Masked Image Modeling

CurMIM: Curriculum Masked Image Modeling

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Liu, Hao Wang, Kun Han, Yudong Wang, Haocong Hu, Yupeng Wang, Chunxiao Nie, Liqiang School of Software Shandong University Jinan China Key Laboratory of Computing Power Network and Information Security Ministry of Education Qilu University of Technology Shandong Academy of Sciences Jinan China School of Computer Science and Technology Harbin Institute of Technology Shenzhen China

ISBN: (纸本)9798350368741

Masked Image Modeling (MIM), following "mask-and-reconstruct" scheme, is a promising self-supervised method to learn scalable visual representation. Studies indicate that selecting an effective mask strategy is vital for MIM. However, existing approaches often rely on static pre-defined priors, which limit their ability to adapt mask strategies dynamically for network optimization. In this paper, we focus on the learning process of the network and introduce human-like curriculum into MIM for dynamic representation refinement, and propose an end-to-end framework Curriculum Masked Image Modeling (CurMIM). CurMIM consists of two components: Mask Priority Measurer, which acts as a curriculum learner to determine mask priority values using the network's intrinsic state information, and Dual Adaptive Selector, which serves as a curriculum scheduler to create effective masks based on these values. With negligible extra parameters, our curriculum-based method consistently establishes noticeable improvements across varying model sizes and benchmarks, showing effectiveness and generalization. © 2025 IEEE.

关键词： Curriculum Learning Image Representation Masked Image Modeling Visual Pre-training

来源：评论

学校读者我要写书评

暂无评论

Functional Testing and Repair for ReRAM-Based Deep-Learning Accelerators

IEEE Design and Test

引用

IEEE Design and Test 2025年第3期42卷 66-73页

作者： Li, Wen Wang, Ying Li, Huawei Li, Xiaowei Zou, Kaiwei Shanxi University School of Computer and Information Technology Shanxi Taiyuan030006 China Institute of Big Data Science and Industry Shanxi University Shanxi Taiyuan030006 China State Key Laboratory of Processors Institute of Computing Technology Chinese Academy of Sciences Beijing100190 China University of Chinese Academy of Sciences Beijing100049 China Tsinghua University Department of Electronic Engineering Beijing100084 China

Editor’s notes: This article offers an overview of the RRAMedy framework, designed to perform online testing, fault diagnosis, and fault mitigation in resistive random-access memory (ReRAM) devices, specifically in the context of ReRAM crossbar-based implementations for AI hardware acceleration. —Jyotika Athavale, Synopsys, USA —Haralampos-G. Stratigopoulos, Sorbonne Université, CNRS, LIP6, France. © 2013 IEEE.

关键词： Accuracy Fault tolerant systems Fault tolerance Deep learning Image edge detection Fault detection Computational modeling Training Reliability Hafnium

来源：评论

学校读者我要写书评

暂无评论

Pronounced advance on typhoon track forecast with global convection-permitting model

引用

Science Bulletin 2025年第9期70卷 1424-1428页

作者： Gu, Jun Zhao, Chun Li, Gudongze Feng, Jiawang Xu, Mingyue Du, Qiuyan Xia, Zihan Li, Yubin Chen, Guanghua Hao, Xiaoyu Chen, Junshi An, Hong Deep Space Exploration Laboratory/School of Earth and Space Sciences/CMA-USTC Laboratory of Fengyun Remote Sensing/State Key Laboratory of Fire Science/Institute of Advanced Interdisciplinary Research on High-Performance Computing Systems and Software University of Science and Technology of China Hefei230026 China Laoshan Laboratory Qingdao266237 China CAS Center for Excellence in Comparative Planetology University of Science and Technology of China Hefei230026 China School of Atmospheric Physics Nanjing University of Information Science and Technology Nanjing210044 China Key Laboratory of Cloud-Precipitation Physics and Severe Storms Institute of Atmospheric Physics Chinese Academy of Sciences Beijing100029 China School of Computer Science and Technology University of Science and Technology of China Hefei230026 China

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：