检索结果-内蒙古大学图书馆

Optimization and parallelization Single Particle Cryo-EM Software RELION with GPU

Jisuanji Yanjiu yu Fazhan/Computer Research and Development 2018年第2期55卷 409-417页

作者： Su, Huayou Wen, Wen Li, Dongsheng Science and Technology on Parallel and Distributed Processing Laboratory National University of Defense Technology Changsha410073 China

Single particle cryo-electron microscopy (cryo-EM) is one of the most important methods of macromolecular structure. RELION (regularized likelihood optimization) is an open-source computer program for the refinement of macromolecular structures by single-particle analysis of cryo-EM data. Due to its easy usage and high quality results, RELION has attracted a lot of attentions from researchers. However, the computation requirement of this program is too huge to solve some large molecular structures with CPU, which harpers the popularization of RELION. In this paper, we characterize the algorithm of RELION and parallelize it with GPU. Firstly, the mathematical theory, computer patterns and performance bottlenecks of RELION are analyzed comprehensively. Then, we optimize the program targeting at fine-grained many-core architecture processor, such as GPU. We propose an efficient multi-level parallel model to utilize the powerful computation capacity of many-core processor. In order to achieve high performance, we reconstruct the data structure for GPU continues memory access. To avoid the limitation of GPU memory size, we implement an adaptive framework. The experimental results show that the proposed GPU based algorithm can achieve good performance. When compared with the CPU implementation, the speedup ratio of the application is more than 36 times, while the speedup ratio of compute-intensive algorithm is about 75X. Moreover, the testing results on multi GPUs show that the GPU based implementation has good scalability. © 2018, science Press. All right reserved.

关键词： Graphics processing unit

来源：评论

学校读者我要写书评

暂无评论

Stochastic extra-gradient based alternating direction methods for graph-guided regularized minimization

引用

Frontiers of Information technology & Electronic Engineering 2018年第6期19卷 755-762页

作者： Qiang LAN Lin-bo QIAO Yi-jie WANG College of Computer National University of Defense Technology National Laboratory for Parallel and Distributed Processing National University of Defense Technology

In this study, we propose and compare stochastic variants of the extra-gradient alternating direction method, named the stochastic extra-gradient alternating direction method with Lagrangian function（SEGL） and the stochastic extra-gradient alternating direction method with augmented Lagrangian function（SEGAL）, to minimize the graph-guided optimization problems, which are composited with two convex objective functions in large scale.A number of important applications in machine learning follow the graph-guided optimization formulation, such as linear regression, logistic regression, Lasso, structured extensions of Lasso, and structured regularized logistic regression. We conduct experiments on fused logistic regression and graph-guided regularized regression. Experimental results on several genres of datasets demonstrate that the proposed algorithm outperforms other competing algorithms, and SEGAL has better performance than SEGL in practical use.

关键词： Stochastic optimization Graph-guided minimization Extra-gradient method Fused logistic regression Graph-guided regularized logistic regression

来源：评论

学校读者我要写书评

暂无评论

Modeling complex relationship paths for knowledge graph completion

Modeling complex relationship paths for knowledge graph comp...

引用

作者： Zeng, Ping Tan, Qingping Meng, Xiankai Zhang, Haoyu Xu, Jianjun College of Computer National University of Defense Technology Changsha China National Key Laboratory for Parallel and Distributed Processing Changsha China

Determining the validity of knowledge triples and filling in the missing entities or relationships in the knowledge graph are the crucial tasks for large-scale knowledge graph completion. So far, the main solutions use machine learning methods to learn the low-dimensional distributed representations of entities and relationships to complete the knowledge graph. Among them, translationmodels obtain excellent performance. However, the proposed translation models do not adequately consider the indirect relationships among entities, affecting the precision of the representation. Based on the long short-term memory neural network and existing translation models, we propose a multiple-module hybrid neural network model called TransP. By modeling the entity paths and their relationship paths, TransP can effectively excavate the indirect relationships among the entities, and thus, improve the quality of knowledge graph completion tasks. Experimental results show that TransP outperforms state-of-the-art models in the entity prediction task, and achieves the comparable performance with previous models in the relationship prediction task. © 2018 The Institute of Electronics, Information and Communication Engineers.

关键词： Knowledge graph

来源：评论

学校读者我要写书评

暂无评论

DCIGAN: A distributed Class-Incremental Learning Method Based on Generative Adversarial Networks

DCIGAN: A Distributed Class-Incremental Learning Method Base...

引用

IEEE International Conference on Big Data and Cloud Computing (BdCloud)

作者： Hongtao Guan Yijie Wang Xingkong Ma Yongmou Li Science and Technology on Parallel and Distributed Laboratory College of Computer National University of Defense Technology Changsha Hunan P.R.China

ISBN: (数字)9781728143286

ISBN: (纸本)9781728143293

Class incremental learning has received wide attention due to its better adaptability to the changing characteristics of online learning. However, data cannot be shared between organizations in the data isolated islands scenario. Existing solutions cannot adapt to incremental classes without aggregating data. In this paper, we propose a distributed class-incremental learning framework, called DCIGAN. It uses GAN generators to store the information of past data and continuously update GAN parameters with new data. In particular, we proposed CIGAN to ensure that the distribution of generated pseudo data is as close as possible to the real data on a single node, which guarantees the accuracy of class-incremental learning. Furthermore, we propose GF, a generators fusion method to integrate local generators of multi-nodes into a new global generator. To evaluate the performance of DCIGAN, we conduct experiments on six datasets under various parameter settings on both two and multi nodes distributed scenarios. Extensive experiments confirm that DCIGAN outperform the general baselines and achieve classification accuracy closes to the method of data aggregating.

关键词： Generators Generative adversarial networks distributed databases Diseases Servers Hospitals Training

来源：评论

学校读者我要写书评

暂无评论

Surveying concurrency bug detectors based on types of detected bugs

引用

science China(Information sciences) 2017年第3期60卷 5-31页

作者： Zhendong WU Kai LU Xiaoping WANG Science and Technology on Parallel and Distributed Processing Laboratory National University of Defense Technology College of Computer National University of Defense Technology

Concurrency bugs widely exist in concurrent programs and have caused severe failures in the real world. Researchers have made significant progress in detecting concurrency bugs, which improves software reliability. In this paper, we survey the most up-to-date and well-known concurrency bug detectors. We categorize the existing detectors based on the types of concurrency bugs. Consequently, we analyze data race detectors, atomicity violation detectors, order violation detectors, and deadlock detectors, respectively. We also discuss some other techniques which are mostly related to concurrency bug detection, including schedule bounding techniques, interleaving optimizing techniques, path expanding techniques, and deterministic replay techniques. Additionally, we statistically analyze the reviewed detectors and get some interesting findings, for instance, nearly 86% of previous detectors focus on data races and atomicity violations, and dynamic approaches are popular(74%). We also discuss the limitations of previous detectors, finding that 91% of previous detectors suffer from false negatives and 64% of previous detectors suffer from runtime overhead. Based on the reviewed detectors and statistical analysis, we conclude some future research directions, including accuracy, performance,applicability, and integrality.

关键词： concurrency bug detection data race atomicity violation order violation deadlock

来源：评论

学校读者我要写书评

暂无评论

Detection of regions with the least impact on true and fake image classification through reinforcement learning

引用

Journal of Physics: Conference Series 2020年第1期1693卷

作者： Wei Li Peng Qiao Yong Dou Science and Technology on Parallel and Distributed Laboratory National University of Defense Technology Changsha 410005 China

With the development of artificial intelligence [1-3], convolutional neural networks (CNNs) have made significant progress in image generation and manipulation. The generated images using facial image synthesis methods, e.g., Faceswap [4], Deepfakes [5], Face2face [6], NeuralTextures [7], are very realistic and even difficult to distinguish by human eyes. It poses serious challenges to social and personal security. There are many methods that can detect synthetic facial images with an accuracy rate of over 99%. But the synthetic facial image classification model can only give a true or fake result, and it is impossible to know which part is the discordant area. Therefore, we propose and experimentally verify a method that using reinforcement learning to select the region in the image that has the least impact on the classification model.

关键词：

来源：评论

学校读者我要写书评

暂无评论

TMDA: Task-Specific Multi-source Domain Adaptation via Clustering Embedded Adversarial Training

TMDA: Task-Specific Multi-source Domain Adaptation via Clust...

引用

IEEE International Conference on Data Mining (ICDM)

作者： Haotian Wang Wenjing Yang Zhipeng Lin Yue Yu State Key Laboratory of High Performance Computing (HPCL) National University of Defense Technology China National Laboratory for Parallel and Distributed Processing National University of Defense Technology China Peng Cheng Laboratory Shenzhen China

Beyond classical domain-specific adversarial training, a recently proposed task-specific framework has achieved a great success in single source domain adaptation by utilizing task-specific decision boundaries. However, compared to single-source-single-target setting, multi-source domain adaptation (MDA) shows more powerful capability to handle with most real-life cases. To align target domain with diverse multi-source domains using task-specific decision boundaries, we provide a deep insight of task-specific framework on MDA for the first time. Accordingly, we propose a novel task-specific multi-source domain adaptation method (TMDA) with a clustering embedded adversarial training process. Specifically, the proposed TMDA detects and refines less discriminative target representations through a max-min optimization over two adversarial task-specific classifiers. Moreover, our analysis implies that scattered multi-source representations disturb the adversarial training under the task-specific framework. To tight up the dispersed source representations, we embeds a relationship-based domain clustering into TMDA. Empirical results demonstrate that our TMDA outperforms state-of-the-art methods on toy dataset, sentiment analysis and digit classification.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Learning-based Depth Estimation aided Visual SLAM Approach

arXiv

引用

arXiv 2019年

作者： Geng, Mingyang Shang, Suning Ding, Bo Wang, Huaimin Zhang, Pengfei Zhang, Lei National Key Laboratory of Parallel and Distributed Processing College of Computer National University of Defense Technology China National Key Laboratory of Integrated Automation of Process Industry Northeastern University China

Existing visual-based SLAM systems mainly utilize the three-dimensional environmental depth information from RGB-D cameras to complete the robotic synchronization localization and map construction task. However, the RGB-D camera maintains a limited range for working and is hard to accurately measure the depth information in a far distance. Besides, the RGB-D camera will easily be influenced by strong lighting and other external factors, which will lead to a poor accuracy on the acquired environmental depth information. Recently, deep learning technologies have achieved great success in the visual SLAM area, which can directly learn high-level features from the visual inputs and improve the estimation accuracy of the depth information. Therefore, deep learning technologies maintain the potential to extend the source of the depth information and improve the performance of the SLAM system. However, the existing deep learning-based methods are mainly supervised and require a large amount of ground-truth depth data, which is hard to acquire because of the realistic constraints. In this paper, we first present an unsupervised learning framework, which not only uses image reconstruction for supervising but also exploits the pose estimation method to enhance the supervised signal and add training constraints for the task of monocular depth and camera motion estimation. Furthermore, we successfully exploit our unsupervised learning framework to assist the traditional ORB-SLAM system when the initialization module of ORB-SLAM method could not match enough features. Qualitative and quantitative experiments have shown that our unsupervised learning framework performs the depth estimation task comparably to the supervised methods and outperforms the previous state-of-the-art approach by 13.5% on KITTI dataset. Besides, our unsupervised learning framework could significantly accelerate the initialization process of ORB-SLAM system and effectively improve the accuracy on environme

关键词： Unsupervised learning

来源：评论

学校读者我要写书评

暂无评论

MixedCache: Enabling Flow Directed Rule-Caching Scheme based on Heterogeneous Cache for OpenFlow

引用

Journal of Physics: Conference Series 2020年第1期1693卷

作者： Jialun Yang Yi Yang Jiahao Liu Rulin Liu Peng Qiao Baosheng Wang College of Computer Science and Technology National University of Defense Technology Changsha 410073 China Science and Technology on Parallel and Distributed Processing Laboratory National University of Defense Technology Changsha 410073 China

OpenFlow switches in SDN use Multiple Flow Tables (MFTs) for fine-grained flow control. Commodity switches integrate hardware storage resources such as SRAM and TCAM to store flow tables to achieve high-speed lookups. Many increased flow tables are rapidly exhausting these hardware storage resources, which makes the switches have to balance high-speed search and massive storage. The rule-caching scheme is a popular method to solve this problem, which caches the most commonly used rules into hardware storage resources. The existing rule-caching schemes are based on single hardware storage resources, and they cannot flexibly adjust the caching strategy according to the traffic characteristics. Simultaneously, the deployed commodity switches face the problem of difficulty in changing the size of SRAM and TCAM. This paper innovatively proposes the MixedCache scheme, which makes full use of the hardware storage resources in the switch according to the skewed characteristics of network traffic. MixedCache stores the large flows in SRAM by exact match and stores the small flows in the TCAM by wildcard match. MixedCache does not need to change the size of the deployed switch hardware storage resources, but makes full use of existing resources. Compared with the rule-caching scheme based on the exact match, the cache hit rate can increase by up to 15.61%. Compared with the rule-caching scheme based on the wildcard match, the cache hit rate can increase by up to 29.69%.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Correction to: A compressed matrix sequence method for solving normal equations of bundle adjustment

引用

Machine Vision and Applications 2021年第4期32卷 1-1页

作者： Peng, Jiaxin Liu, Jie Wei, Hua Laboratory of Software Engineering for Complex Systems School of Computer Science National University of Defense Technology Changsha China Parallel and Distributed Processing Laboratory School of Computer Science National University of Defense Technology Changsha China Algorithm Lab Changsha Morale Network Technology Co. Ltd. Changsha China Water and Land Resources Research Center of the Middle Reaches of Yangtze River School of Resources Environment Science and Engineering Hubei University of Science and Technology Xianning China

A correction to this paper has been published: https://***/10.1007/s00138-021-01222-5

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：