检索结果-内蒙古大学图书馆

Proceedings of the 23rd USENIX Conference on File and Storage Technologies

作者： Zhuohui Duan Hao Feng Haikun Liu Xiaofei Liao Hai Jin Bangyu Li National Engineering Research Center for Big Data Technology and System Service Computing Technology and System Lab/Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology China

ISBN: (纸本)9781939133458

The key-value separation is renowned for its significant mitigation of the write amplification inherent in traditional LSM trees. However, KV separation potentially increases performance overhead in the management of Value region, especially for garbage collection (GC) operation that is used to reduce the redundant space occupation. In response, many efforts have been made to optimize the GC mechanism for KV separation. However, our analysis indicates that such solution based on trade-offs between CPU and I/O overheads cannot simultaneously satisfy the three requirements of KV separated systems in terms of throughput, tail latency, and space usage. This limitation hinders their real-world *** this paper, we introduce AegonKV, a "three-birds-one-stone" solution that comprehensively enhances the throughput, tail latency, and space usage of KV separated systems. AegonKV first proposes a SmartSSD-based GC offloading mechanism to enable asynchronous GC operations without competing with LSM read/write for bandwidth or CPU. AegonKV leverages offload-friendly data structures and hardware/ software execution logic to address the challenges of GC offloading. Experiments demonstrate that AegonKV achieves the largest throughput improvement of 1.28-3.3 times, a significant reduction of 37%-66% in tail latency, and 15%-85% in space overhead compared to existing KV separated systems.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Expediting Distributed GNN Training with Feature-only Partition and Optimized Communication Planning

Expediting Distributed GNN Training with Feature-only Partit...

引用

IEEE Annual Joint Conference: INFOCOM, IEEE computer and Communications Societies

作者： Bingqian Du Jun Liu Ziyue Luo Chuan Wu Qiankun Zhang Hai Jin 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 University of Science and Technology Wuhan China Department of Electrical and Computer Engineering The Ohio State University USA Department of Computer Science The University of Hong Kong Hong Kong School of Cyber Science and Engineering Huazhong University of Science and Technology Wuhan China

ISBN: (数字)9798350383508

ISBN: (纸本)9798350383515

Feature-only partition of large graph data in distributed Graph Neural Network (GNN) training offers advantages over commonly adopted graph structure partition, such as minimal graph preprocessing cost and elimination of cross-worker subgraph sampling burdens. Nonetheless, performance bottleneck of GNN training with feature-only partitions still largely lies in the substantial communication overhead due to cross-worker feature fetching. To reduce the communication overhead and expedite distributed training, we first investigate and answer two key questions on convergence behaviors of GNN model in feature-partition based distribute GNN training: 1) As no worker holds a complete copy of each feature, can gradient exchange among workers compensate for the information loss due to incomplete local features? 2) If the answer to the first question is negative, is feature fetching in every training iteration of the GNN model necessary to ensure model convergence? Based on our theoretical findings on these questions, we derive an optimal communication plan that decides the frequency for feature fetching during the training process, taking into account bandwidth levels among workers and striking a balance between model loss and training time. Extensive evaluation demonstrates consistent results with our theoretical analysis, and the effectiveness of our proposed design.

关键词： Training Time-frequency analysis Costs Computational modeling Distributed databases Bandwidth Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Discrete Geometric Coded Data Layout for Large-scale Object Storage systems 21

Discrete Geometric Coded Data Layout for Large-scale Object ...

引用

21st IEEE International Symposium on Parallel and Distributed Processing with Applications, 13th IEEE International Conference on Big Data and Cloud Computing, 16th IEEE International Conference on Social Computing and Networking and 13th International Conference on Sustainable Computing and Communications, ISPA/BDCloud/SocialCom/SustainCom 2023

作者： Tian, Yi Xu, Guangping Yang, Hongzhang Ni, Yue Cao, JiaXin Yang, Lei Tianjin University of Technology School of Computer Science and Engineering Tianjin China Novel Software Technology Tianjin Key Laboratory of Intelligence Computing Tianjin China Key Laboratory of Computer Vision and System Ministry of Education China Information Technology Co. Ltd. China

ISBN: (纸本)9798350329223

Regenerating codes are new network codes proposed to reduce the data required for fault repair, which can improve the recovery efficiency of faulty nodes in data storage systems. However, unlike Reed-Solomon code, which repairs at the granularity of bytes, regenerating codes require data stored in large chunks but leads to severe read amplification, which reads out excess data when degraded read objects and increases degraded read time. That reflects a mutual constraint between improving recovery efficiency and degraded read performance, as manifested in the amplification of data reads, a vital issue considered in this *** solve this problem, we propose a new type of data layout - discrete geometric, which splits the object into a series of geometric sequences of data blocks. They are placed discretely into corresponding containers on the disks at different nodes, with containers of the same size made into a strip for encoding. The discrete characteristic ensures lower repair costs for degraded reads. The geometric characteristic ensures the repair performance of regenerating codes by large blocks, and read amplification can be mitigated through small blocks. To reduce IOPS for discrete geometric, we propose Discrete Geometric-Locally Regenerating Codes (DG-LRCs), guaranteeing lower degraded read latency while improving recovery *** results show that the degraded read time of DG-LRCs compared to regenerating codes combined with geometric partitioning is 22.56% lower at 2Gbpsand 60.56% lower at 4Gbps, and the recovery performance is 7.04 times better than that of RS code. © 2023 IEEE.

关键词： Efficiency

来源：评论

学校读者我要写书评

暂无评论

Securely Fine-tuning Pre-trained Encoders Against Adversarial Examples

Securely Fine-tuning Pre-trained Encoders Against Adversaria...

引用

IEEE Symposium on Security and Privacy

作者： Ziqi Zhou Minghui Li Wei Liu Shengshan Hu Yechao Zhang Wei Wan Lulu Xue Leo Yu Zhang Dezhong Yao Hai Jin 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 University of Science and Technology School of Software Engineering Huazhong University of Science and Technology Hubei Engineering Research Center on Big Data Security Hubei Key Laboratory of Distributed System Security School of Cyber Science and Engineering Huazhong University of Science and Technology School of Information and Communication Technology Griffith University

ISBN: (数字)9798350331301

ISBN: (纸本)9798350331318

With the evolution of self-supervised learning, the pre-training paradigm has emerged as a predominant solution within the deep learning landscape. Model providers furnish pre-trained encoders designed to function as versatile feature extractors, enabling downstream users to harness the benefits of expansive models with minimal effort through fine-tuning. Nevertheless, recent works have exposed a vulnerability in pre-trained encoders, highlighting their susceptibility to downstream-agnostic adversarial examples (DAEs) meticulously crafted by attackers. The lingering question pertains to the feasibility of fortifying the robustness of downstream models against DAEs, particularly in scenarios where the pre-trained encoders are publicly accessible to the *** this paper, we initially delve into existing defensive mechanisms against adversarial examples within the pre-training paradigm. Our findings reveal that the failure of current defenses stems from the domain shift between pre-training data and downstream tasks, as well as the sensitivity of encoder parameters. In response to these challenges, we propose Genetic Evolution-Nurtured Adversarial Fine-tuning (Gen-AF), a two-stage adversarial fine-tuning approach aimed at enhancing the robustness of downstream models. Gen-AF employs a genetic-directed dual-track adversarial fine-tuning strategy in its first stage to effectively inherit the pre-trained encoder. This involves optimizing the pre-trained encoder and classifier separately while incorporating genetic regularization to preserve the model’s topology. In the second stage, Gen-AF assesses the robust sensitivity of each layer and creates a dictionary, based on which the top-k robust redundant layers are selected with the remaining layers held fixed. Upon this foundation, we conduct evolutionary adaptability fine-tuning to further enhance the model’s generalizability. Our extensive experiments, conducted across ten self-supervised training methods and six d

关键词： Training Adaptation models Sensitivity Accuracy Self-supervised learning Genetics Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Attention-enhanced Relation Network for Few-shot Image Classification 23

Attention-enhanced Relation Network for Few-shot Image Class...

引用

6th International Conference on Image and Graphics Processing, ICIGP 2023

作者： Li, Jinyang Tong, Jiahui Gao, Guangyu Xu, Wenbin School of Computer Science and Technology Beijing Institute of Technology China Science and Technology on Complex System Control and Intelligent Agent Cooperation Laboratory Beijing Electro-Mechanical Engineering Institute Beijing China Science and Technology on Optical Radiation Laboratory Beijing Institute of Environmental Characteristics Beijing100854 China

ISBN: (纸本)9781450398572

Traditional deep learning models firmly rely on a large amount of labeled data during pre-training. Whereas it lacks generalization in the face of unfamiliar categories. Recently, few-shot learning is a hot topic in computer vision to classify unseen classes with limited labels. A representative approach is to extract features from the support and query sets, respectively, and compare similarities via metric learning. However, convolutional neural networks often focus only on a local region and ignore the global region, which severely reduces the accuracy of the matching. Specifically, in this paper, we pile lightweight attention-based blocks in the embedding module, which combines an adaptive kernel size 2D convolutional network with a cross-channel attention mechanism to encode multi-scale features and implicitly increase the receptive field. The SE-relation module chooses to construct learnable non-linear comparators to compare the relationship utilizing channel information. Finally, we show experimental results on standard few-shot testing benchmarks such as mini-ImageNet and tiered-ImageNet to demonstrate effectiveness. © 2023 ACM.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

GLC++: Source-Free Universal Domain Adaptation through Global-Local Clustering and Contrastive Affinity Learning

arXiv

引用

arXiv 2024年

作者： Qu, Sanqing Zou, Tianpei Röhrbein, Florian Lu, Cewu Chen, Guang Tao, Dacheng Jiang, Changjun The School of Automotive Engineering Department of Computer Science Tongji University China The Faculty of Computer Science Chemnitz University of Technology Germany The Department of Computer Science Shanghai Jiao Tong University China The School of Computer Science Faculty of Engineering The University of Sydney Australia The Department of Computer Science Key Laboratory of Embedded System and Service Computing Ministry of Education Tongji University China

Deep neural networks often exhibit sub-optimal performance under covariate and category shifts. Source-Free Domain Adaptation (SFDA) presents a promising solution to this dilemma, yet most SFDA approaches are restricted to closed-set scenarios. In this paper, we explore Source-Free Universal Domain Adaptation (SF-UniDA) aiming to accurately classify "known" data belonging to common categories and segregate them from target-private "unknown" data. We propose a novel Global and Local Clustering (GLC) technique, which comprises an adaptive one-vs-all global clustering algorithm to discern between target classes, complemented by a local k-NN clustering strategy to mitigate negative transfer. Despite the effectiveness, the inherent closed-set source architecture leads to uniform treatment of "unknown" data, impeding the identification of distinct "unknown" categories. To address this, we evolve GLC to GLC++, integrating a contrastive affinity learning strategy. We examine the superiority of GLC and GLC++ across multiple benchmarks and category shift scenarios. Remarkably, in the most challenging open-partial-set scenarios, GLC and GLC++ surpass GATE by 16.7% and 18.6% in H-score on VisDA, respectively. GLC++ enhances the novel category clustering accuracy of GLC by 4.3% in open-set scenarios on Office-Home. Furthermore, the introduced contrastive learning strategy not only enhances GLC but also significantly facilitates existing methodologies. Copyright © 2024, The Authors. All rights reserved.

关键词： Clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

How to Select Pre-Trained Code Models for Reuse? A Learning Perspective

arXiv

引用

arXiv 2025年

作者： Bi, Zhangqian Wan, Yao Chu, Zhaoyang Hu, Yufei Zhang, Junyi Zhang, Hongyu Xu, Guandong Jin, Hai National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab Wuhan China School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China School of Big Data and Software Engineering Chongqing University Chongqing China School of Computer Science University of Technology Sydney Sydney Australia

Pre-training a language model and then fine-tuning it has shown to be an efficient and effective technique for a wide range of code intelligence tasks, such as code generation, code summarization, and vulnerability detection. However, pre-training language models on a large-scale code corpus is computationally expensive. Fortunately, many off-the-shelf Pre-trained Code Models (PCMs), such as CodeBERT, CodeT5, CodeGen, and Code Llama, have been released publicly. These models acquire general code understanding and generation capability during pre-training, which enhances their performance on downstream code intelligence tasks. With an increasing number of these public pre-trained models, selecting the most suitable one to reuse for a specific task is essential. In this paper, we systematically investigate the reusability of PCMs. We first explore three intuitive model selection methods that select by size, training data, or brute-force fine-tuning. Experimental results show that these straightforward techniques either perform poorly or suffer high costs. Motivated by these findings, we explore learning-based model selection strategies that utilize pre-trained models without altering their parameters. Specifically, we train proxy models to gauge the performance of pre-trained models, and measure the distribution deviation between a model's latent features and the task's labels, using their closeness as an indicator of model transferability. We conduct experiments on 100 widely-used open-source PCMs for code intelligence tasks, with sizes ranging from 42.5 million to 3 billion parameters. The results demonstrate that learning-based selection methods reduce selection time to 100 seconds, compared to 2,700 hours with brute-force fine-tuning, with less than 6% performance degradation across related tasks. © 2025, CC BY-NC-SA.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Single Block Encoder-Decoder Transformer Model for Multi-Step Traffic Flow Forecasting 6

Single Block Encoder-Decoder Transformer Model for Multi-Ste...

引用

6th IEEE Symposium on computers and Informatics, ISCI 2024

作者： Omar, Mas Yakub, Fitri Talip, Mohamad Sofian Abu Maslan, Mohd Nazmin Sinha, Vijay Kumar Muljono, Muljono Universiti Teknologi Malaysia Wee Laboratory Malaysia-Japan International Institute of Technology Kuala Lumpur Malaysia Universiti Teknologi Malaysia Malaysia-Japan International Institute of Technology Department of Electronic System Engineering Kuala Lumpur Malaysia University of Malaya Faculty of Engineering Department of Electrical Engineering Kuala Lumpur Malaysia Universiti Teknikal Malaysia Faculty of Industrial and Manufacturing Technology and Engineering Melaka Malaysia Chitkara University School of Engineering & Technology Department of Computer Science & Engineering Himachal Pradesh India Universitas Dian Nuswantoro Faculty of Computer Science Semarang Indonesia

ISBN: (纸本)9798350353853

Accurate traffic flow forecasting is crucial for managing and planning urban transportation systems. Despite the widespread use of sequence modelling models like Long Short-Term Memory (LSTM) for this purpose, the potential of Transformer models remains underexplored. This is particularly true for the simplest form of a single block encoder-decoder Transformer model, which can be finely tuned through optimised hyperparameters. This paper examines the performance of a singular horizon-step forecasting method for multi-step traffic flow forecasting using a proposed Single Block Encoder-Decoder Transformer model optimised with a Grid Search algorithm. Results demonstrate that this model can enhance forecasting accuracy compared to the state-of-the-art LSTM model typically used for multi-step forecasting. The model effectively captures long-range temporal dependencies within a single road traffic flow dataset. It was tested on hourly traffic flow data to forecast the next 24 hours for the I5-North freeway in California, sourced from the Caltrans Performance Measurement system. The optimal configuration included an embedding dimension of 32, a feed-forward dimension of 128, and 8 attention heads. Results show a significant improvement, with a 4.7% reduction in Root Mean Squared Error compared to an LSTM model with two hidden layers of 100 neurons each, showcasing the potential of Single Block Encoder-Decoder Transformer models for real-world traffic prediction applications. © 2024 IEEE.

关键词： Mean square error

来源：评论

学校读者我要写书评

暂无评论

A Comparative Study of YOLOv5 and YOLOv8 for Helmet Detection Inside ATMs

A Comparative Study of YOLOv5 and YOLOv8 for Helmet Detectio...

引用

Advancement in Computation & computer Technologies (InCACCT), International Conference on

作者： Prashant Kumar Sharma Lakshmipriya B Poonguzhali N Jayalakshmy S School of Computer Science & Engineering Vellore Institute of Technology Chennai India School of Engineering Jawaharlal Nehru University New Delhi India School of Computer & System Sciences Jawaharlal Nehru University New Delhi India Electronics & Communication Engineering IFET College of Engineering Villupuram India

ISBN: (数字)9798331543891

ISBN: (纸本)9798331543907

Automated helmet detection inside automated teller machine (ATM) cabin for cash dispense can enhance security as well as serves as a means to prevent heist and vandalism. This study presents a comparative study on helmet detection inside ATMs using two prominent object detection you look only once (YOLO) v5 and YOLOv8. The system facilitates detection of helmets from closed circuit television (CCTV) footages to help in identifying the people entering into the ATM by wearing helmets and also in avoiding potential intrusion. In addition to precision and recall, mean average precision (mAP), a standard object detection metric to measure the accuracy of localization is used in this paper to assess the performance of helmet detection models constructed using YOLOv5 and YOLOv8. Experimental evaluation carried out in this study elucidates that YOLOv8 performs better than YOLOv5. By integrating the proposed model with IoT techniques, this model improves the safety and security. In addition, the incorporation of other alert models with this system offers a safer environment for the users inside the ATM cabin.

关键词： YOLO Training Performance evaluation Location awareness Head TV Automated teller machine Safety Security Standards

来源：评论

学校读者我要写书评

暂无评论

How to Select Pre-Trained Code Models for Reuse? A Learning Perspective

How to Select Pre-Trained Code Models for Reuse? A Learning ...

引用

IEEE International Conference on Software Analysis, Evolution and Reengineering (SANER)

作者： Zhangqian Bi Yao Wan Zhaoyang Chu Yufei Hu Junyi Zhang Hongyu Zhang Guandong Xu Hai Jin National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab Wuhan China School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China School of Big Data and Software Engineering Chongqing University Chongqing China School of Computer Science University of Technology Sydney Sydney Australia

ISBN: (数字)9798331535100

ISBN: (纸本)9798331535117

Pre-training a language model and then fine-tuning it has shown to be an efficient and effective technique for a wide range of code intelligence tasks, such as code generation, code summarization, and vulnerability detection. However, pre-training language models on a large-scale code corpus is compu-tationally expensive. Fortunately, many off-the-shelf Pre-trained Code Models (PCMs), such as CodeBERT, CodeT5, CodeGen, and Code Llama, have been released publicly. These models acquire general code understanding and generation capability during pre-training, which enhances their performance on downstream code intelligence tasks. With an increasing number of these public pre-trained models, selecting the most suitable one to reuse for a specific task is essential. In this paper, we systematically investigate the reusability of PCMs. We first explore three intuitive model selection methods that select by size, training data, or brute-force fine-tuning. Experimental results show that these straightforward techniques either perform poorly or suffer high costs. Motivated by these findings, we explore learning-based model selection strategies that utilize pre-trained models without altering their parameters. Specifically, we train proxy models to gauge the performance of pre-trained models, and measure the distribution deviation between a model's latent features and the task's labels, using their closeness as an indicator of model transferability. We conduct experiments on 100 widely-used open-source PCMs for code intelligence tasks, with sizes ranging from 42.5 million to 3 billion parameters. The results demonstrate that learning-based selection methods reduce selection time to 100 seconds, compared to 2,700 hours with brute-force fine-tuning, with less than 6% performance degradation across related tasks.

关键词： Phase change materials Analytical models Adaptation models Codes Costs Computational modeling Training data Machine learning Software Software development management

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：