检索结果-内蒙古大学图书馆

10th IEEE Joint International Information Technology and Artificial Intelligence Conference, ITAIC 2022

作者： Liu, Yang Lu, Jun College of Computer Science and Technology Heilongjiang University Harbin150080 China Heilongjiang University Key Laboratory of Database and Parallel Computing of Heilongjiang Province Harbin150080 China

ISBN: (数字)9781665422079

ISBN: (纸本)9781665422079

Neural architecture search aims to automatically design network architectures by machines, which is expected to bring about a new revolution in machine learning. Despite high expectations, the effectiveness and efficiency of existing neural architecture search solutions are unclear, and the inefficiency of neural architecture search solutions is mainly attributed to inaccurate architecture evaluation. In this work, this paper divides the large search space of neural architecture search into multiple modules, effectively reducing the error caused by shared parameters to ensure that the candidate architectures in the modules are fully trained. By adopting a block search method, this paper can Evaluate all candidate architectures in a module, addressing the problem of inaccurate evaluation. In addition, high-quality data is necessary. Under the guidance of the teacher model, combined with the ground truth, the neural architecture search is supervised by the double loss function. Under the constraint of the same amount of computation, the method in this paper achieves 77.3% top-level accuracy on ImageNet., which is 1.0% higher than the efficient network EfficientNet-B0, and 0.3% higher than the DNA method without additional computation. Therefore, the double loss block neural architecture search model proposed in this paper significantly improves the effectiveness of neural architecture search. © 2022 IEEE.

关键词： Distillation

来源：评论

学校读者我要写书评

暂无评论

Research on combinatorial encryption method

Research on combinatorial encryption method

引用

2014 International Conference on Information Science, Electronics and Electrical Engineering, ISEEE 2014

作者： Lu, Jun Mo, Juan Lv, Xingfeng Zhang, Zhuo College of Computer Science and Technology Heilongjiang University College of CS HLJU Harbin China Key Laboratory of Database and Parallel Computing Heilongjiang Province KLDPC HLJU Harbin China

ISBN: (纸本)9781479931965

Combinatorial encryption algorithm is advanced. Combinatorial encryption method makes use of the variety of ordinal in universal combination code. Firstly, the theory of universal combinatorial coding is expatiated. Secondly, researches are done in combinatorial encryption characteristic, key generating method, encryption and decryption algorithms. At last, key space is analyzed. Combinatorial encryption technology breaks through the thought of encryption. It is based on combinatorial theory and the code element is adopted as the basic unit in encryption computing. Furthermore, the key generating method is unique. Combinatorial encryption technology makes a new attempt to cipher system. © 2014 IEEE.

关键词： Cryptography

来源：评论

学校读者我要写书评

暂无评论

Node classification problem based on potential information and representation learning between nodes 7

Node classification problem based on potential information a...

引用

7th International Conference on Information Science and Control Engineering, ICISCE 2020

作者： Li, Mingming Yang, Yan Heilongjiang University School of Computer Science Technology Hearbin Province China Key Laboratory of Database and Parallel Computing of Heilongjiang Province Harbin Province China

ISBN: (纸本)9781728164069

Representation learning of social network nodes refers to the feature expression of nodes in low-dimensional vector space. Most of the current methods are based on the network topology to achieve node feature learning, but the potential information of the network topology is not fully utilized. To solve this problem, we propose a graph embedding algorithm based on the similarity between nodes. This method obtains a new network topology through the second-order similarity between nodes. This network topology can enhance the correlation between nodes, thereby improving the effect of node classification in the network. We use Macro-F1 and Micro-F1 evaluation indicators to evaluate the benchmark algorithm and the SNE algorithm proposed in this paper on four real data sets. The experimental results show that the SNE algorithm can improve the effect of node classification. © 2020 IEEE.

关键词： Network topology

来源：评论

学校读者我要写书评

暂无评论

Cross-Modality Encoder Representations Based On External Attention Mechanism 3

Cross-Modality Encoder Representations Based On External Att...

引用

3rd International Conference on Neural Networks, Information and Communication Engineering, NNICE 2023

作者： Zheng, Yudong Lu, Jun College of Computer Science and Technology Heilongjiang University Harbin China College of Computer Science and Technology Heilongjiang University Key Laboratory of Database and Parallel Computing of Heilongjiang Province Harbin China

ISBN: (纸本)9798350335972

With the prevalence of deep learning, people use multi-modality information for interpretation and reasoning. In this paper, a cross-modality encoder CMEEA (cross-modality encoder representation based on external attention mechanism) is adopted to improve the accuracy of the model. The two external storage units of CMEEA can be regarded as a dictionary of the whole dataset to improve the performance of the network, while learning more representative features of the input and reducing the computational cost. This paper uses five pretraining tasks that help the model learn internal modal and cross-modal relationships. It also verifies the generalizability of the model by applying the pretraining cross-modal model to the visual inference task NLVR2, and improves the previous results by 0.1%. By fine-tuning the pretrain parameters, the model improves by 1.3% on visual question answering (VQA). © 2023 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

An implementation of the QR iterations for finding eigenvalues of matrices with CUDA on GPU

An implementation of the QR iterations for finding eigenvalu...

引用

2013 International Conference on Mechatronic Sciences, Electric Engineering and Computer, MEC 2013

作者： Ren, Meirui Ren, Meihui Zhang, Wei Ping Wang, Tao Tian, Ning Li, Jinbao Guo, Longjiang School of Computer Science and Technology Heilongjiang University Harbin150080 China Harbin Research Institute of Electrical Instrumentation Harbin150028 China Key Laboratory of Database and Parallel Computing of Heilongjiang Province Harbin150080 China

ISBN: (纸本)9781479925650

Based on the principle of QR loop iterations, this paper implements a parallel algorithm based on the hardware of GPU (Graphic Process Unit) by using routines from CUDA (Computer Unified Device Architecture) to find the eigenvalues of general matrices. CPU and GPU computing.card form a client-server computing.framework. Here, CPU can be regarded as a client and GPU card is considered as a computing.server. For the experiment environment, this paper chooses a GPU card with the model of NVIDIA GeForce GTX460 as a server side and a CPU with the model of Intel Core is-760 quad-core as a client side. Win7 64-bit is selected as the operating system. The parallel implementation consists of two parts: PA-H and PA-QR. PA-H is a procedure that transforms a matrix A into the Hessenberg matrix B. PA-QR is the actual parallel algorithm of the QR iterations that is imposed on the Hessenberg matrix B for finding eigenvalues of the general matrix A. The speedup ratio of the proposed algorithm is jarless when the number of iterations becomes greater. The experimental results show that the parallel implementation with CUDA on GPU only makes use of less running time than traditional sequential algorithms. The speedup ratio of PA-H is between 1.79 and 7.81. The speedup ratio of PA-QR is between 3.24 and 118.9. Especially, when the order of general matrix equals 8192, the amount of iterations becomes 10000. The speedup ratio of the PA-H and the PA-QR can run up to 7.81 and 118.9 respectively. © 2013 IEEE.

关键词： Matrix algebra

来源：评论

学校读者我要写书评

暂无评论

The parallel processing for promoter data base on OpenMP

The parallel processing for promoter data base on OpenMP

引用

International Conference on Electronic Engineering and Information Science, ICEEIS 2015

作者： Shi, Y. Lu, J. Zheng, J.J. College of Computer Science and Technology Heilongjiang University China College of Computer Science and Technology Heilongjiang University Key Laboratory of Database and Parallel Computing of Heilongjiang Province Harbin China

ISBN: (纸本)9781138027725

The promoters of Arabidopsis and the Poplar are parallel processed and analyzed in this paper. The mainly works include three parts: First, The Motif data are parallel matched with both plant promoter data based on OpenMP. Then, the matching result is processed by frequent mining technology. The experimental result shows that the parallel processing time based on OpenMP is far less than the serial time. Second, to increase the calculation accuracy, the algorithm of solving the P value is improved by the prime split method. At last, according to the frequent process result, the shared frequent items of the two plants are analyzed. © 2015 Taylor & Francis Group, London.

关键词： Application programming interfaces (API)

来源：评论

学校读者我要写书评

暂无评论

DMSA: Dynamic Multi-Scale Unsupervised Semantic Segmentation Based On Adaptive Affinity 48

DMSA: Dynamic Multi-Scale Unsupervised Semantic Segmentation...

引用

48th IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2023

作者： Yang, Kun Lu, Jun Heilongjiang University College of Computer Science and Technology Harbin150080 China Jiaxiang Industrial Technology Research Institute of Hlju Jining China Key Laboratory of Database and Parallel Computing of Heilongjiang Province Harbin China

ISBN: (纸本)9781728163277

The proposed method in this paper proposes an end-to-end unsupervised semantic segmentation architecture DMSA based on four loss functions. The framework uses Atrous Spatial Pyramid Pooling (ASPP) module to enhance feature extraction. At the same time, a dynamic dilation strategy is designed to better capture multi-scale context information. Secondly, a Pixel-Adaptive Refinement (PAR) module is introduced, which can adaptively refine the initial pseudo labels after feature fusion to obtain high quality pseudo labels. Experiments show that the proposed DSMA framework is superior to the existing methods on the saliency dataset. On the COCO 80 dataset, the MIoU is improved by 2.0, and the accuracy is improved by 5.39. On the Pascal VOC 2012 Augmented dataset, the MIoU is improved by 4.9, and the accuracy is improved by 3.4. In addition, the convergence speed of the model is also greatly improved after the introduction of the PAR module. © 2023 IEEE.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Minimum latency data aggregation scheduling in MR-MC wireless sensor networks

引用

Tongxin Xuebao/Journal on Communications 2014年第10期35卷 192-199页

作者： Li, Jin-Bao Wang, Meng Guo, Long-Jiang School of Computer Science and Technology Heilongjiang University Harbin150080 China Key Laboratory of Database and Parallel Computing of Heilongjiang Province Harbin150080 China

The problem of minimum latency data aggregation scheduling in single-radio single-channel wireless sensor networks is a NPC problem and many solutions have been proposed. In multi-radio multi-channel networks, the latency may be lower as a node may receive transmission data from other different nodes simultaneously. It formulates the minimum latency data aggregation scheduling into an optimization problem with joint consider of tree structure, time slot assignment, channel assignment and radio assignment. The problem is divided into constructing data aggregation tree and designing a node schedule subproblems. For each subproblem, heuristic algorithm has been proposed. The simulation results show that the proposed algorithm achieve good performance.

关键词： Scheduling

来源：评论

学校读者我要写书评

暂无评论

Density and distance based probabilistic broadcasting algorithm in mobile sensor networks

引用

Jisuanji Yanjiu yu Fazhan/Computer Research and Development 2014年第1期51卷 151-160页

作者： Shen, Yue Guo, Longjiang Li, Jinbao School of Computer Science and Technology Heilongjiang University Harbin 150080 China Key Laboratory of Database and Parallel Computing Heilongjiang Province Harbin 150001 China

Broadcasting is the most basic manner to transmit messages in mobile wireless sensor networks. However, existing relevant broadcasting algorithms in mobile wireless sensor networks need a large number of intermediate forwarding nodes. The large number of intermediate forwarding nodes causes a mass of redundant message packages. Massive redundant messages will consume much energy and lead to a short life period of mobile wireless sensor networks. This paper proposes a broadcasting algorithm named node density and distance-based probability (NDDP for short). The average forwarding ratio of the algorithm is only 5S/(Nπr2), where S is the area of the networks, N is the amount of nodes in the networks, and r is communication range. Then the average receipt ratio of NDDP is more than 95 percent in theoretical analysis and more than 92 percent in the ns-2 simulation results. Denser the network is, the more energy saving the algorithm is. Simulation results also show that NDDP outperforms the two algorithms in Smite and Sidewinder not only on the aspect of stability but also on energy conservation.

关键词： Energy conservation

来源：评论

学校读者我要写书评

暂无评论

Data collection using wavelet-segment constant compression in wireless sensor networks

引用

Yi Qi Yi Biao Xue Bao/Chinese Journal of Scientific Instrument 2013年第1期34卷 119-127页

作者： Li, Yang Guo, Longjiang Li, Jinbao Ren, Meirui School of Computer Science and Technology Heilongjiang University Harbin 150080 China Key Laboratory of Database and Parallel Computing Heilongjiang Province Harbin 150080 China

In structural health monitoring system, it usually needs to arrange lots of sensor nodes. In order to get the monitoring data in real time, frequent data collection is essential will consume so much energy, especially when the data is multi-dimensional. To reduce communication messages in the frequent data collection applications and as far as possible to extend the network lifetime. This paper proposes a data collection algorithm based on wavelet-segment constant compression(W-SDC), applies compressive sampling theory, and compress the messages with W-SDC before data transmitted, thereby this algorithm reduces the amount of data transmission. When the network is large-scale, it could save communication costs greatly, and extend the network lifetime. The total messages of traditional data collection method is O(N2), compressive data gathering method is O(MN), while the total communication costs of this algorithm is O(mN), where m

关键词： Sensor nodes

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：