检索结果-内蒙古大学图书馆

IEEE International Conference on Systems, Man and Cybernetics

作者： MingLe Zhou ChangLe Yi HongLin Wan Min Li Gang Li DeLong Han Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center (National Supercomputer Center in Jinan) Qilu University of Technology (Shandong Academy of Sciences) Jinan China Shandong Provincial Key Laboratory of Computer Networks Shandong Fundamental Research Center for Computer Science Jinan China College of Physics and Electronic Science Shandong Normal University Jinan China

The detection of defects in printed circuit boards requires high accuracy and realtime performance. Existing industrial detection models generally adopt a pure convolutional structure for ease of deployment. However, the detection accuracy of these models is often insufficient to meet the requirements of the scene. To improve the detection model accuracy and ease of deployment, this paper proposes a convolutional merging Transformer network(CMTRNet). The CMTR-Net model proposes a backbone network (CNN-Former) that uses convolutional modules to replace self-attention, combining the Transformer architecture with a convolutional structure. This approach not only avoids the drawback of self-attention high computation complexity that is detrimental to deployment but also improves the model detection accuracy. Based on CNN-Former, this paper also proposes a feature fusion module that can better fuse the features extracted by CNN-Former. Further-more, based on the CMTRNet model and the characteristics of circuit board defects, this paper proposes a loss function called Melt-IoU, which can make the initial training phase smoother and further improve detection accuracy. Experiments have shown that CMTRNet outperforms existing advanced models on both datasets.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Pulsar candidate recognition based on autoencoder and self-normalizing neural networks

Pulsar candidate recognition based on autoencoder and self-n...

引用

IEEE Joint International information technology and Artificial Intelligence Conference (ITAIC)

作者： Zhihao Wang Peng Fu Meng Guo Key Laboratory of Computing Power Network and Information Security of Ministry of Education Shandong Computer Science Center (National Supercomputing Center in Jinan) (Shandong Academy of Sciences) Qilu University of Technology Jinan Shandong P. R. China Shandong Provincial Key Laboratory of Computer Networks Shandong Fundamental Research Center for Computer Science Jinan Shandong P. R. China

The identification of pulsar candidates is a crucial step in radio astronomy research. With the continuous improvement of modern radio telescope equipment and the increasing scale of pulsar sky survey, a pulsar survey will produce a large number of pulsar candidates. Identifying pulse signals from a large number of pulsar candidates is a prerequisite for discovering new pulsar. In the face of massive pulsar candidate data, it is a huge workload to identify pulsar candidates manually only by experts in relevant fields. In order to improve the efficiency of identifying pulsar candidates, a stacked autoencoder and self-normalizing neural network based pulsar candidate recognition framework AESNN is proposed, which does not need to consider the imbalance of the number of positive and negative samples in the pulsar data set. The pulsar candidate contains many features, but not all of them are important. The stacked autoencoder is used to extract the feature representation of the data, perform unsupervised dimensionality reduction, and input these features into the self-normalizing neural network for training. AESNN introduces self-normalizing properties by introducing 'scaled exponential linear units', and experiments show that AESNN has a high F1 score, precision, and recall, and does not require the use of complex data augmentation methods, making it more advantageous compared to other algorithms.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Whole-proteome based phylogenetic tree construction with inter-amino-acid distances and the conditional geometric distribution profiles

引用

Molecular phylogenetics and evolution 2015年 89卷 37-45页

作者： Xian-Hua Xie Zu-Guo Yu Guo-Sheng Han Wei-Feng Yang Vo Anh Hunan Key Laboratory for Computation and Simulation in Science and Engineering Key Laboratory of Intelligent Computing and Information Processing of Ministry of Education Xiangtan University Hunan 411105 PR China School of Mathematics and Computer Science Gannan Normal University Jiangxi 341000 PR China. Electronic address: xxianhua@***. School of Mathematical Sciences Queensland University of Technology GPO Box 2434 Brisbane QLD 4001 Australia. Electronic address: yuzuguo@***. Hunan Key Laboratory for Computation and Simulation in Science and Engineering Key Laboratory of Intelligent Computing and Information Processing of Ministry of Education Xiangtan University Hunan 411105 PR China. Electronic address: korea10282003@***. Hunan Key Laboratory for Computation and Simulation in Science and Engineering Key Laboratory of Intelligent Computing and Information Processing of Ministry of Education Xiangtan University Hunan 411105 PR China. Electronic address: yangweifeng09@***. School of Mathematical Sciences Queensland University of Technology GPO Box 2434 Brisbane QLD 4001 Australia. Electronic address: v.anh@qut.edu.au.

There has been a growing interest in alignment-free methods for whole genome comparison and phylogenomic studies. In this study, we propose an alignment-free method for phylogenetic tree construction using whole-proteome sequences. Based on the inter-amino-acid distances, we first convert the whole-proteome sequences into inter-amino-acid distance vectors, which are called observed inter-amino-acid distance profiles. Then, we propose to use conditional geometric distribution profiles (the distributions of sequences where the amino acids are placed randomly and independently) as the reference distribution profiles. Last the relative deviation between the observed and reference distribution profiles is used to define a simple metric that reflects the phylogenetic relationships between whole-proteome sequences of different organisms. We name our method inter-amino-acid distances and conditional geometric distribution profiles (IAGDP). We evaluate our method on two data sets: the benchmark dataset including 29 genomes used in previous published papers, and another one including 67 mammal genomes. Our results demonstrate that the new method is useful and efficient.

关键词： Alignment-free whole-proteome comparison Conditional geometric distribution profile Inter-amino-acid distances Phylogenetic tree

来源：评论

学校读者我要写书评

暂无评论

CMC v2: Towards More Accurate COVID-19 Detection with Discriminative Video Priors

arXiv

引用

arXiv 2022年

作者： Hou, Junlin Xu, Jilan Zhang, Nan Wang, Yi Zhang, Yuejie Zhang, Xiaobo Feng, Rui School of Computer Science Shanghai Key Laboratory of Intelligent Information Processing Fudan University China Academy for Engineering and Technology Fudan University China Shanghai AI Laboratory China Children’s Hospital of Fudan University National Children’s Medical Center Shanghai China

This paper presents our solution for the 2nd COVID-19 Competition, occurring in the framework of the AIMIA Workshop at the European Conference on computer Vision (ECCV 2022). In our approach, we employ the winning solution last year which uses a strong 3D Contrastive Mixup Classification network (CMC v1) as the baseline method, composed of contrastive representation learning and mixup classification. In this paper, we propose CMC v2 by introducing natural video priors to COVID-19 diagnosis. Specifically, we adapt a pre-trained (on video dataset) video transformer backbone to COVID-19 detection. Moreover, advanced training strategies, including hybrid mixup and cutmix, slice-level augmentation, and small resolution training are also utilized to boost the robustness and the generalization ability of the model. Among 14 participating teams, CMC v2 ranked 1st in the 2nd COVID-19 Competition with an average Macro F1 Score of 89.11%. © 2022, CC BY.

关键词： COVID-19

来源：评论

学校读者我要写书评

暂无评论

Feature Aggregation Network for Memory-Based Industrial Anomaly Detection

Feature Aggregation Network for Memory-Based Industrial Anom...

引用

IEEE International Conference on Data Science in Cyberspace (DSC)

作者： Min Li Delong Han Rui Shao Jinghui He Faculty of Data Science City University of Macau China Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center (National Supercomputer Center in Jinan) Shandong Computer Science Center (National Supercomputer Center in Jinan) Qilu University of Technology (Shandong Academy of Sciences) Jinan China Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center (National Supercomputer Center in Jinan) Qilu University of Technology (Shandong Academy of Sciences) Jinan China Shandong Provincial Key Laboratory of Computer Networks Shandong Fundamental Research Center for Computer Science Jinan China

ISBN: (数字)9798350391367

ISBN: (纸本)9798350391374

In the process of large-scale industrial production, complex defects of different shapes and sizes can occur in industrial products and their components, and these defects seriously affect the performance of industrial products and their components, and being able to quickly and accurately identify defective industrial products and their components is an important task. Memory-based methods are usually used for anomaly detection, and these methods often rely on pre-trained models for feature extraction, but have high requirements on feature quality. In addition, patch-based methods mainly focus on local differences in feature embedding when detecting anomalies without explicitly exploiting broader contextual information, thus greatly affecting the robustness and accuracy of anomaly detection. Aiming at the above problems and requirements, this paper proposes a Feature Aggregation Network for Memory-Based Industrial Anomaly Detection (FANet), which is based on the pre-extraction of anomalous features, attentional attention to key features, information aggregation of multi-level and multi-scale features, and the FANet provides advanced performance in industrial image anomaly detection and localisation by pre-extraction of anomalous features, attentional attention to key features, information aggregation of multi-layered and multi-scale features, and introduction of environmental information in the anomaly score computation to perform industrial anomaly detection tasks, and using core set subsampling to reduce storage requirements. On the MVTec AD, FANet has an image-level AUROC score of 99.1% and a pixel-level AUROC score of 98.2%. Similarly high detection results were achieved on the ViSA dataset, further demonstrating the robustness of FANet.

关键词： Training Runtime Shape Memory management Training data Production Feature extraction Robustness Anomaly detection Testing

来源：评论

学校读者我要写书评

暂无评论

A Multi-Path Orchestration Method for Cloud-Edge Data Transmission Based on DRL

A Multi-Path Orchestration Method for Cloud-Edge Data Transm...

引用

International Conference on Parallel and Distributed Systems (ICPADS)

作者： Zhihui Wang Xubing Dou Shuyun Li Huan Ma Xinchang Zhang Maoli Wang Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center (National Supercomputer Center in Jinan) Qilu University of Technology (Shandong Academy of Sciences) Jinan China Shandong Provincial Key Laboratory of Computer Networks Shandong Fundamental Research Center for Computer Science Jinan China Shandong Normal University Jinan China QuFu normal university Jining China

ISBN: (数字)9798331515966

ISBN: (纸本)9798331515973

In addressing the multi-path problem of cloud-edge data transmission, this paper introduces a multi-path orchestration method based on deep reinforcement learning (DRL) within the context of software-defined networks (SDN), termed Deep Q-Routing (DQR). This method establishes optimal data transmission paths between the cloud and multiple edges, based on available bandwidth and transmission delay, while avoiding path congestion to improve bandwidth utilization and reduce transmission delay. The optimal routing in SDN is defined through the design of state space sets, reward functions, and neural networks. Simulation experiments were conducted to evaluate the DQR method. Compared to existing routing methods, the proposed DQR method, which includes a reward function with weighted parameters, demonstrates superior overall network performance.

关键词： Cloud computing Scalability Neural networks Distributed databases Bandwidth Routing Deep reinforcement learning Delays Data communication Software defined networking

来源：评论

学校读者我要写书评

暂无评论

Prompt-Based Few-Shot Text Classification with Virtual Class Representation

Prompt-Based Few-Shot Text Classification with Virtual Class...

引用

International Conference on computer and Communications (ICCC)

作者： Zhengyu Ju Zhao Li Ye Li Shiwei Wu Canjun Wang Shandong Academy of Sciences Qilu University of Technology Jinan China Key Laboratory of Computing Power Network and Information Security Ministry of Education Jinan China Shandong Computer Science Center National Supercomputer Center in Jinan Jinan China Shandong Provincial Key Laboratory of Computer Networks Jinan China Shandong Fundamental Research Center for Computer Science Jinan China Evay Info Jinan China

Over the past few years, "Pre-training + Fine-tuning" has been highly successful in Natural Language processing (NLP) tasks, but it has limitations in few-shot tasks and zero-shot tasks. Recently, "Prompt-based Tuning + Pretrained Language models (PLMs)" has achieved great success in few-shot stage and zero-shot stage, including few-shot text classification. Prompt-based tuning wraps suggestive characters and [MASK] tokens into a raw input sequence. Additionally, Prompt learning uses a verbalizer to map the semantic information of the [MASK] position to the final classification prediction. The verbalizer contains manual verbalizer and automatic verbalizer, etc. However, manual verbalizers are too dependent on human experience and external knowledge, while automatic verbalizers still face significant challenges. In this work, we introduce a novel approach for constructing a verbalizer termed as “virtual class representation”. This method entails assigning a fixed-length sequence of digits to represent each label in the dataset. We train a small-scale randomly initialized embedding mechanism to achieve preliminary semantic information for each class when virtual class representation is introduced into the embedding mechanism. The semantic information of each category is then further extracted by a convolution operation, and the final model regression value is obtained by calculating the similarity between the last hidden state of PLM and the semantic vector of the class. Our extensive experiments demonstrate the effectiveness of the proposed method on few-shot text classification tasks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Ekblom Promoting Adaptive Algorithm for System Identification

TechRxiv

引用

TechRxiv 2022年

作者： Huang, Xinqi Li, Yingsong Han, Xiao Tu, Huawei College of Information and Communication Engineering Harbin Engineering University Harbin150001 China Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Anhui University Anhui Hefei China College of Computer Science and Technology Harbin Engineering University Harbin150001 China Department of Computer Science and Information Technology La Trobe University MelbourneVIC3086 Australia

We propose an Ekblom promoting adaptive algorithm (EPAA) which uses Ekblom norm to construct a data reusing scheme to achieve better performance under impulsive noise (IN) environments. By exerting Ekblom-norm constraint on the a posterior error vector, a sliding window cost function is created to realize the EPAA that is derived to reduce IN effects. The derivation and theoretical convergence analysis of the EPAA are presented in detail. The simulation examples are setup to illustrate the superior performance of the EPAA compared to popular data reusing algorithms for system identification under IN environments. © 2022, CC BY.

关键词： Adaptive filtering

来源：评论

学校读者我要写书评

暂无评论

Effect of WC addition on microstructure and properties of powder metallurgy CoCrNi medium entropy alloy

引用

Materials Today Communications 2023年 36卷

作者： Huang, Jianhong He, Jichang Li, Zhanjiang Shen, Li Chen, Li Chang, Fa Dai, Pinqiang Tang, Qunhua College of Materials Science and Engineering Fujian University of Technology Fuzhou350118 China Fujian Provincial Key Laboratory of Advanced Materials Processing and Application Fuzhou350118 China College of Materials Science and Engineering Fuzhou University Fuzhou350116 China School of Mechanical Electrical & Information Engineering Putian University Putian351100 China

To surpass the current level of mechanical properties exhibited by CoCrNi medium-entropy alloys (MEA), (CoCrNi)100−x(WC)x (x = 0, 2, 4 and 6 for molar ratio) MEA samples were prepared using powder metallurgy with atomized powder as the raw material, and the influence of WC content on the microstructural evolution and mechanical properties of the sintered MEAs was investigated. As the WC content increased, the phase structure of the sintered MEAs changed from a single FCC phase to an FCC+Cr23C6 phase, while the grain size of the matrix decreased. Consequently, the strength of the alloy exhibited significant improvement. For instance, the tensile yield and fracture strengths of (CoCrNi)98(WC)2 were measured at 574 and 971 MPa, respectively, while maintaining a plasticity of approximately 41.6%. Furthermore, the addition of WC led to a considerable increase in the hardness for (CoCrNi)94(WC)6, reaching 314.7 HV, which was nearly 50% higher than that of the CoCrNi matrix. The increased hardness proved beneficial in enhancing the material's wear resistance, which first increased and then decreased. The wear mechanisms primarily involved abrasive and oxidation wears. The observed increase in yield strength can be attributed to the combined effects of multiple strengthening mechanisms. The individual contributions of these mechanisms were quantitatively evaluated and compared to the actual values, demonstrating consistency with the calculated results. Notably, solid-solution strengthening emerged as the principle strengthening mechanism. © 2023 Elsevier Ltd

关键词： Chromium alloys

来源：评论

学校读者我要写书评

暂无评论

Sparse belief propagation algorithm based on latent dirichlet allocation

引用

Diangong Jishu Xuebao/Transactions of China Electrotechnical Society 2015年 30卷 506-511页

作者： Tu, Xionggang Chen, Jun Yang, Lu Yan, Jianfeng School of Design and Arts Zhejiang Industry Polytechnic College ShaoxingZhejiang312000 China School of Computer Science and Technology Soochow University Suzhou215006 China Provincial Key Laboratory for Computer Information Processing Technology Soochow University Suzhou215006 China

Latent Dirichlet allocation(LDA) is a popular and unsupervised tool for reducing dimension, has been applied in text mining and information retrieval. Belief propagation is competitive in both speed and accuracy compared with other inference algorithms of LDA. In this paper, we propose belief propagation algorithm with sparseness constraints and use L1-norm, L2-norm to measure the sparseness of vector. In iteration, we project the topic-word distribution to a sparse space for getting better semantic representation. Experimental results show that sparse belief propagation algorithm improves the accuracy of text clustering and classification than belief propagation, Gibbs sampling and nonnegative matrix factorization. ©, 2015, The editorial office of Transaction of China Electrotechnical Society. All right reserved.

关键词： Belief propagation

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：