检索结果-内蒙古大学图书馆

23rd IEEE International Conference on High Performance Computing and Communications, 7th IEEE International Conference on data Science and Systems, 19th IEEE International Conference on Smart City and 7th IEEE International Conference on Dependability in Sensor, Cloud and big data Systems and Applications, HPCC-DSS-SmartCity-DependSys 2021

作者： Zhang, Zhan Yue, Jianhui Liao, Xiaofei Jin, Hai School of Computer Science and Technology Huazhong University of Science and Technology National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab Wuhan China Michigan Technological University Computer Science Department HoughtonMI United States

ISBN: (纸本)9781665494571

Write-ahead log and data encryption technologies are employed to ensure both crash consistency and data security for persistent memory (PM). The encryption/decryption of both data and log requests increase the memory request latency, degrading the system performance. To tackle this issue, this paper proposes a novel log-aware memory encryption scheme to reduce encryption/decryption operations, without compromising data security. Specifically, redo log blocks are stored as ciphertext for their corresponding data blocks to avoid encryption/decryption of log data blocks when in-place update operations are applied to them. We further design a compact log record layout with fewer encryption metadata in log records to reduce logging traffic. Our simulation results show that the transaction throughput of the proposed design outperforms the baseline design and the state-of-the-art design by 87.3% and 55.3% on average, respectively. © 2021 IEEE.

关键词： Smart cities System performance Simulation Memory management Layout Memory architecture Metadata

来源：评论

学校读者我要写书评

暂无评论

Reveal training performance mystery between Tensor Flow and PyTorch in the single GPU environment

引用

Science China(Information Sciences) 2022年第1期65卷 147-163页

作者： Hulin DAI Xuan PENG Xuanhua SHI Ligang HE Qian XIONG Hai JIN National Engineering Research Center for Big Data Technology and System Service Computing Technology and System LabSchool of Computer Science and Technology Huazhong University of Science and Technology Department of Computer Science University of Warwick

Deep learning has gained tremendous success in various fields while training deep neural networks(DNNs) is very compute-intensive, which results in numerous deep learning frameworks that aim to offer better usability and higher performance to deep learning practitioners. Tensor Flow and Py Torch are the two most popular frameworks. Tensor Flow is more promising within the industry context, while Py Torch is more appealing in academia. However, these two frameworks differ much owing to the opposite design philosophy:static vs dynamic computation graph. Tensor Flow is regarded as being more performance-friendly as it has more opportunities to perform optimizations with the full view of the computation graph. However, there are also claims that Py Torch is faster than Tensor Flow sometimes, which confuses the end-users on the choice between them. In this paper, we carry out the analytical and experimental analysis to unravel the mystery of comparison in training speed on single-GPU between Tensor Flow and Py Torch. To ensure that our investigation is as comprehensive as possible, we carefully select seven popular neural networks, which cover computer vision, speech recognition, and natural language processing(NLP). The contributions of this work are two-fold. First, we conduct the detailed benchmarking experiments on Tensor Flow and Py Torch and analyze the reasons for their performance difference. This work provides the guidance for the end-users to choose between these two frameworks. Second, we identify some key factors that affect the performance,which can direct the end-users to write their models more efficiently.

关键词： deep learning performance comparison TensorFlow PyTorch

来源：评论

学校读者我要写书评

暂无评论

Capsule: An Out-of-Core Training Mechanism for Colossal GNNs

引用

Proceedings of the ACM on Management of data 2025年第1期3卷 1-30页

作者： Yongan Xiang Zezhong Ding Rui Guo Shangyou Wang Xike Xie S. Kevin Zhou School of Computer Science and Technology University of Science and Technology of China (USTC) Hefei Anhui China & Data Darkness Lab Suzhou Institute for Advanced Research USTC Suzhou Jiangsu China School of Artificial Intelligence and Data Science USTC Hefei Anhui China & Data Darkness Lab Suzhou Institute for Advanced Research USTC Suzhou Jiangsu China School of Biomedical Engineering USTC Suzhou Jiangsu China & Data Darkness Lab Suzhou Institute for Advanced Research USTC Suzhou Jiangsu China School of Biomedical Engineering USTC Suzhou Jiangsu China & MIRACLE Center Suzhou Institute for Advanced Research USTC Suzhou Jiangsu China

Cutting-edge platforms of graph neural networks (GNNs), such as DGL and PyG, harness the parallel processing power of GPUs to extract structural information from graph data, achieving state-of-the-art (SOTA) performance in fields such as recommendation systems, knowledge graphs, and bioinformatics. Despite the computational advantages provided by GPUs, these GNN platforms struggle with scalability challenges due to the colossal graphical structures processed and the limited memory capacities of GPUs. In response, this work introduces Capsule, a new out-of-core mechanism for large-scale GNN training. Unlike existing out-of-core GNN systems, which use main or secondary memory as operative memory and use CPU kernels during non-backpropagation computation, Capsule uses GPU memory and GPU kernels. By substantially leveraging the parallelization capabilities of GPUs, Capsule significantly enhances GNN training efficiency. In addition, Capsule can be smoothly integrated to mainstream open-source GNN frameworks, DGL and PyG, in a play-and-plug manner. Through a prototype implementation and comprehensive experiments on real datasets, we demonstrate that Capsule can achieve up to a 12.02× improvement in runtime efficiency, while using only 22.24% of the main memory, compared to SOTA out-of-core GNN systems.

关键词： gpus graph data management graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Backdoor Defense via Deconfounded Representation Learning

Backdoor Defense via Deconfounded Representation Learning

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Zaixi Zhang Qi Liu Zhicai Wang Zepu Lu Qingyong Hu Anhui Province Key Lab of Big Data Analysis and Application School of Computer Science and Technology University of Science and Technology of China State Key Laboratory of Cognitive Intelligence Hefei Anhui China University of Science and Technology of China Hong Kong University of Science and Technology

Deep neural networks (DNNs) are recently shown to be vulnerable to backdoor attacks, where attackers embed hidden backdoors in the DNN model by injecting a few poisoned examples into the training dataset. While extensive efforts have been made to detect and remove backdoors from backdoored DNNs, it is still not clear whether a backdoor-free clean model can be directly obtained from poisoned datasets. In this paper, we first construct a causal graph to model the generation process of poisoned data and find that the backdoor attack acts as the confounder, which brings spurious associations between the input images and target labels, making the model predictions less reliable. Inspired by the causal understanding, we propose the Causality-inspired Backdoor Defense (CBD), to learn deconfounded representations for reliable classification. Specifically, a backdoored model is intentionally trained to capture the confounding effects. The other clean model dedicates to capturing the desired causal effects by minimizing the mutual information with the confounding representations from the backdoored model and employing a sample-wise re-weighting scheme. Extensive experiments on multiple benchmark datasets against 6 state-of-the-art attacks verify that our proposed defense method is effective in reducing backdoor threats while maintaining high accuracy in predicting benign samples. Further analysis shows that CBD can also resist potential adaptive attacks. The code is available at https://***/zaixizhang/CBD.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Enabling Efficient Large Recommendation Model Training with Near CXL Memory Processing

Enabling Efficient Large Recommendation Model Training with ...

引用

Annual International Symposium on Computer Architecture, ISCA

作者： Haifeng Liu Long Zheng Yu Huang Jingyi Zhou Chaoqiang Liu Runze Wang Xiaofei Liaot Hai Jinf Jingling Xue National Engineering Research Center for Big Data Technology and System/Services Computing Technology and System Lab/Cluster and Grid Computing Lab Huazhong University of Science and Technology China School of Computer Science and Engineering University of New South Wales Australia Zhejiang Lab Hangzhou China

ISBN: (数字)9798350326581

ISBN: (纸本)9798350326598

Personalized recommendation systems have become one of the most important Internet services nowadays. A critical challenge of training and deploying the recommendation models is their high memory capacity and bandwidth demands, with the embedding layers occupying hundreds of GBs to TBs of storage. The advent of memory disaggregation technology and Compute Express Link (CXL) provides a promising solution for memory capacity scaling. However, relocating memory-intensive embedding layers to CXL memory incurs noticeable performance degradation due to its limited transmission bandwidth, which is significantly lower than the host memory bandwidth. To address this, we introduce ReCXL, a CXL memory disaggregation system that utilizes near-memory processing for scalable, efficient recommendation model training. ReCXL features a unified, hardwareefficient NMP architecture that processes the entire embedding training within CXL memory, minimizing data transfers over the bandwidth-limited CXL and enhancing internal bandwidth. To further improve the performance, ReCXL incorporates softwarehardware co-optimizations, including sophisticated dependencyfree prefetching and fine-grained update scheduling, to maximize hardware utilization. Evaluation results show that ReCXL outperforms the CPU-GPU baseline and the naïve CXL memory by $7.1 \times \sim 10.6 \times(9.4 \times$ on average) and $12.7 \times \sim 31.3 \times(22.6 \times$ on average), respectively.

关键词： Training Degradation Scalability Prefetching Memory management Web and internet services Bandwidth

来源：评论

学校读者我要写书评

暂无评论

New entanglement-assisted quantum codes from negacyclic codes

arXiv

引用

arXiv 2023年

作者： Chen, Xiaojing Lu, Xingbo Zhu, Shixin Jiang, Wan Wang, Xindi School of Internet Anhui University Anhui Hefei230039 China School of Mathematics Hefei University of Technology Anhui Hefei230601 China Key Laboratory of Knowledge Engineering with Big Data Ministry of Education Hefei University of Technology Anhui Hefei230601 China School of Computer and Information Hefei University of Technology Anhui Hefei230601 China

The theory of entanglement-assisted quantum error-correcting codes (EAQECCs) is a generalization of the standard stabilizer quantum error-correcting codes, which can be possibly constructed from any classical codes by relaxing the duality condition and utilizing preshared entanglement between the sender and receiver. In this paper, a new family of EAQECCs is constructed from negacyclic codes of length n = q2a+1, where q is an odd prime power, a = m22+1 and m is an odd integer. Some new entanglement-assisted quantum maximum distance separable (EAQMDS) codes are obtained in the sense that their parameters are not covered by the previously known ones. Copyright © 2023, The Authors. All rights reserved.

关键词： Quantum entanglement

来源：评论

学校读者我要写书评

暂无评论

3D Landmark Detection on Human Point Clouds: A Benchmark and A Dual Cascade Point Transformer Framework

arXiv

引用

arXiv 2024年

作者： Zhang, Fan Mao, Shuyi Li, Qing Peng, Xiaojiang College of Big Data and Internet Shenzhen Technology University Shenzhen China Department of Electrical and Computer Engineering Georgia Institute of Technology Shenzhen China AI Lab Lenovo Research Shenzhen China

3D landmark detection plays a pivotal role in various applications such as 3D registration, pose estimation, and virtual try-on. While considerable success has been achieved in 2D human landmark detection or pose estimation, there is a notable scarcity of reported works on landmark detection in unordered 3D point clouds. This paper introduces a novel challenge, namely 3D landmark detection on human point clouds, presenting two primary contributions. Firstly, we establish a comprehensive human point cloud dataset, named HPoint103, designed to support the 3D landmark detection community. This dataset comprises 103 human point clouds created with commercial software and actors, each manually annotated with 11 stable landmarks. Secondly, we propose a Dual Cascade Point Transformer (D-CPT) model for precise point-based landmark detection. D-CPT gradually refines the landmarks through cascade Transformer decoder layers across the entire point cloud stream, simultaneously enhancing landmark coordinates with a RefineNet over local regions. Comparative evaluations with popular point-based methods on HPoint103 and the public dataset DHP19 demonstrate the dramatic outperformance of our D-CPT. Additionally, the integration of our RefineNet into existing methods consistently improves performance. Code and data will be released soon. Copyright © 2024, 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

KubeSPT: Stateful Pod Teleportation for Service Resilience with Live Migration

引用

IEEE Transactions on Services Computing 2025年

作者： Zhang, Hansheng Wu, Song Fan, Hao Huang, Zhuo Xue, Weibin Yu, Chen Ibrahim, Shadi Jin, Hai Huazhong University of Science and Technology National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong China

Container orchestration systems, such as Kubernetes, streamline containerized application deployment. As more and more applications are being deployed in Kubernetes, there is an increasing need for rescheduling - relocating a running pod to different nodes - due to system upgrades, node failures, and load-balancing optimizations. Live migration, which transfers services from source nodes to target nodes with minimal downtime, is the ideal support for rescheduling. However, implementing live migration for pods that run stateful services is challenging, because Kubernetes manages pods as stateless. First, the current pod's network namespace initialization process causes a mismatch in the network state between the migrated pod and internal containers. Second, migrating the memory state results in extended downtime. Third, Kubernetes operations on pods do not consider preserving the state of the pods. Therefore, we propose KubeSPT to achieve live migration of stateful pods in rescheduling scenarios. Firstly, we synchronize the network state of pods and internal containers by controlling packet flow and implement fast service redirection. Secondly, we introduce a Hot data and Lazy-Restore method for memory restoration to reduce migration downtime. Finally, we decouple pod migration operations from other Kubernetes operations to ensure compatibility with live migration. Experimental results show that KubeSPT reduces downtime by 86%-93% compared to current rescheduling methods. © 2008-2012 IEEE.

关键词： Restoration

来源：评论

学校读者我要写书评

暂无评论

An Anchor-Free Refining Feature Pyramid Network for Dense and Multioriented Wheat Spikes Detection under UAV

引用

IEEE Transactions on Instrumentation and Measurement 2025年 74卷

作者： Jiao, Lin Liu, Haiyun Liang, Zheng Chen, Peng Wang, Rujing Liu, Kang Anhui University National Engineering Research Center for Agro-Ecological Big Data Analysis and Application School of Internet Hefei230601 China Hefei Institutes of Physical Science Chinese Academy of Sciences Institute of Intelligent Machines Hefei230031 China University of Science and Technology of China Science Island Branch Graduate School Anhui Hefei230031 China Anhui University School of Internet Hefei230031 China School of Internet Institutes of Physical Science and Information Technology Anhui University National Engineering Research Center for Agro-Ecological Big Data Analysis and Application Anhui Hefei230601 China Institute of Intelligent Machines Hefei Institutes of Physical Science Chinese Academy of Science Hefei230031 China Hong Kong Polytechnic University Department of Aeronautical and Aviation Engineering Hong Kong

The number of wheat spikes is a crucial index for evaluating the yield, and the precise detection of wheat spikes in an image plays an important role. Among various methods, deep learning-based approaches show impressive results in the task of wheat spike detection. However, the precise detection and recognition of wheat spike encounters large challenges due to complicated backgrounds, arbitrary orientations, and dense distribution in wheat spike images. To alleviate these issues, we have developed an anchor-free refining feature pyramid network (AFRFPN) that gets rid of horizontal bounding boxes (HBBs) from the network. First, the refining feature pyramid network (RFPN) has been introduced into extract richer features of wheat spike with highly variant appearances and multiple scales. Then, learning from the idea of coarse-to-fine, the two-stage anchor-free oriented detection (AFOD) module has been designed. The AFOD module first generates a set of coarse detection (CoDet) results in the way of anchor-free and then further fines them to achieve high-quality predicting bounding boxes (BBs). The number of wheat spike images is insufficient, resulting in poor performance of wheat spike detection modules. To mitigate the lack of the data in the task of oriented wheat spike detection, based on the global wheat head detection (GWHD) dataset, we released a new large-scale wheat spike dataset by relabeling the samples, termed it as rotated GWHD (R-GWHD) dataset. Massive experiments show that the proposed method can achieve 90.6% mAP and 96.7% recall, outperforming other state-of-the-art methods. Additionally, the experiments related to the counting of wheat spikes have been conducted, showing that the developed module can achieve the MAE of 4.95 and RMSE of 7.68, which demonstrates the excellent performance of the proposed method. © 2024 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

FlexiFed: Personalized Federated Learning for Edge Clients with Heterogeneous Model Architectures 23

FlexiFed: Personalized Federated Learning for Edge Clients w...

引用

2023 World Wide Web Conference, WWW 2023

作者： Wang, Kaibin He, Qiang Chen, Feifei Chen, Chunyang Huang, Faliang Jin, Hai Yang, Yun School of Computer Science and Technology Huazhong University of Science and Technology China Department of Computing Technologies Swinburne University of Technology Australia School of Information Technology Deakin University Australia Faculty of Information Technology Monash University Australia Guangxi Key Lab of Human-machine Interaction and Intelligent Decision Nanning Normal University China National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab Huazhong University of Science and Technology Wuhan430074 China

ISBN: (纸本)9781450394161

Mobile and Web-of-Things (WoT) devices at the network edge account for more than half of the world's web traffic, making a great data source for various machine learning (ML) applications, particularly federated learning (FL) which offers a promising solution to privacy-preserving ML feeding on these data. FL allows edge mobile and WoT devices to train a shared global ML model under the orchestration of a central parameter server. In the real world, due to resource heterogeneity, these edge devices often train different versions of models (e.g., VGG-16 and VGG-19) or different ML models (e.g., VGG and ResNet) for the same ML task (e.g., computer vision and speech recognition). Existing FL schemes have assumed that participating edge devices share a common model architecture, and thus cannot facilitate FL across edge devices with heterogeneous ML model architectures. We explored this architecture heterogeneity challenge and found that FL can and should accommodate these edge devices to improve model accuracy and accelerate model training. This paper presents our findings and FlexiFed, a novel scheme for FL across edge devices with heterogeneous model architectures, and three model aggregation strategies for accommodating architecture heterogeneity under FlexiFed. Experiments with four widely-used ML models on four public datasets demonstrate 1) the usefulness of FlexiFed;and 2) that compared with the state-of-the-art FL scheme, FlexiFed improves model accuracy by 2.6%-9.7% and accelerates model convergence by 1.24 × -4.04 ×. © 2023 ACM.

关键词： Speech recognition

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：