检索结果-内蒙古大学图书馆

Collaborative Bi-Aggregation for Directed Graph Embedding

学校读者我要写书评

暂无评论

SSRN 2023年

作者： Liu, Linsong Chen, Ke-Jia Liu, Zheng Jiangsu Key Laboratory of Big Data Security & Intelligent Processing Nanjing University of Posts and Telecommunications Jiangsu Nanjing210023 China School of Computer Science Nanjing University of Posts and Telecommunications Jiangsu Nanjing210023 China State Key Laboratory of Novel Software Technology Nanjing University Jiangsu Nanjing210093 China

Directed graph is able to model asymmetric relationships between nodes and research on directed graph embedding is of great significance in downstream graph analysis and inference. Learning source and target embeddings of nodes separately to preserve edge asymmetry has become the dominant approach, but also poses challenge for learning representations of low or even zero in/out degree nodes that are ubiquitous in sparse graphs. In this paper, a collaborative bi-directional aggregation method (COBA) for directed graph embedding is proposed. Firstly, the source and target embeddings of the central node are learned by aggregating from the counterparts of the source and target neighbors, respectively;Secondly, the source/target embeddings of the zero in/out degree central nodes are enhanced by aggregating the counterparts of opposite-directional neighbors (i.e. target/source neighbors);Finally, source and target embeddings of the same node are correlated to achieve collaborative aggregation. Both the feasibility and rationality of the model are theoretically analyzed. Extensive experiments on real-world datasets demonstrate that COBA comprehensively outperforms state-of-the-art methods on multiple tasks and meanwhile validates the effectiveness of proposed aggregation strategies. © 2023, The Authors. All rights reserved.

关键词： Directed graphs

SLBERT: An Improved FLAT-BERT for Named Entity Recognition

学校读者我要写书评

暂无评论

SSRN 2022年

作者： Fei, Yuefan Xu, Xiaolong Jiangsu Key Laboratory of Big Data Security & Intelligent Processing Nanjing University of Posts and Telecommunications Jiangsu Nanjing China School of Computer Science Nanjing University of Posts and Telecommunications Jiangsu Nanjing China

Named Entity Recognition (NER) aims to identify named entities mentioned in unstructured text and classify them into predefined named entity classes. Although models using word fusion have achieved relatively high performance, the recognition of ambiguous entities in text remains challenging, and the performance of the models depends on a sufficient amount of manually labeled data;exploring models that fuse more feature dimensions and schemes that fuse existing datasets is of farreaching interest. In this work, we propose an improved scheme of FLAT-BERT, SLBERT, that can be used in the field of named entity *** scheme consists of two phases that more fully exploit the global semantics of the text. In the first stage, the model is trained based on the current dataset, another dataset is predicted for cross-dataset fusion, and the correctness of label annotation is improved by entity confidence modeling. In the second stage, the global semantics of the text is extracted using Sentence BERT and embedded into the feature matrix encoded using Lattice BERT for subsequent Token classification. We have conducted experiments on several datasets, and the experimental results demonstrate the effectiveness of the proposed scheme in this paper. In terms of F1 values, compared with FLAT-BERT, the improvement is 0.15%, 0.34%, 0.61%, and 0.58% on datasets OntoNotes, MSRA, Resume, and Weibo, respectively. © 2022, The Authors. All rights reserved.

关键词： Semantics

A Point Cloud Registration Method for Substations Based on an Improved SAC-IA Algorithm

学校读者我要写书评

暂无评论

A Point Cloud Registration Method for Substations Based on a...

Chinese Automation Congress (CAC)

作者： Lei Wang Yiping Chen Haonan Zong Lulu Wang Jun Yin School of Computer Nanjing University of Posts and Telecommunications Nanjing PR China Jiangsu Key Laboratory of Big Data Security and Intelligent Processing Nanjing PR China School of Internet of Things Nanjing University of Posts and Telecommunications Nanjing PR China

This paper proposes a point cloud registration method for substations based on an improved SAC-IA algorithm. This method optimizes the SAC-IA algorithm by filtering the randomly selected point pairs to ensure that the triangles formed by the selected points in the model and scene point clouds have certain similarities. Furthermore, the registration error judgment of the SAC-IA algorithm has been improved. In addition to using the Euclidean distance between the registered model and scene point clouds as the judgment standard, this method also considers the SHOT difference and average curvature difference. Successful registration is only considered when all three standards are satisfied. The experimental research shows that the time performance of this method is twice as high as that of the traditional SAC-IA algorithm on average, and the recognition accuracy is improved in the recognition results.

关键词：

A comprehensive survey on shadow removal from document images: datasets, methods, and opportunities

学校读者我要写书评

暂无评论

Vicinagearth 2025年第1期2卷 1-18页

作者： Wang, Bingshu Li, Changping Zou, Wenbin Zhang, Yongjun Chen, Xuhang Chen, C.L. Philip School of Software Northwestern Polytechnical University Xi’an China Guangdong Provincial Key Laboratory of Intelligent Information Processing & Shenzhen Key Laboratory of Media Security Shenzhen University Shenzhen China Guangdong Key Laboratory of Intelligent Information Processing College of Electronics and Information Engineering Shenzhen University Shenzhen China Yongjun Zhang is with the State Key Laboratory of Public Big Data College of Computer Science and Technology Guizhou University Guiyang China School of Computer Science and Engineering Huizhou University Huizhou China School of Computer Science and Engineering South China University of Technology and Pazhou Lab Guangzhou China

With the rapid development of document digitization, people have become accustomed to capturing and processing documents using electronic devices such as smartphones. However, the captured document images often suffer from issues like shadows and noise due to environmental factors, which can affect their readability. To improve the quality of captured document images, researchers have proposed a series of models or frameworks and applied them in distinct scenarios such as image enhancement, and document information extraction. In this paper, we primarily focus on shadow removal methods and open-source datasets. We concentrate on recent advancements in this area, first organizing and analyzing nine available datasets. Then, the methods are categorized into conventional methods and neural network-based methods. Conventional methods use manually designed features and include shadow map-based approaches and illumination-based approaches. Neural network-based methods automatically generate features from data and are divided into single-stage approaches and multi-stage approaches. We detail representative algorithms and briefly describe some typical techniques. Finally, we analyze and discuss experimental results, identifying the limitations of datasets and methods. Future research directions are discussed, and nine suggestions for shadow removal from document images are proposed. To our knowledge, this is the first survey of shadow removal methods and related datasets from document images.

关键词：

Raft Consensus Algorithm Based on Credit Model in Consortium Blockchain

学校读者我要写书评

暂无评论

Wuhan University Journal of Natural Sciences 2020年第2期25卷 146-154页

作者： CHEN Yunfang LIU Ping ZHANG Wei School of Computer Science Nanjing University of Posts and TelecommunicationsNanjing 210023JiangsuChina Jiangsu Key Laboratory of Big Data Security and Intelligent Processing Nanjing University of Posts and TelecommunicationsNanjing 210023JiangsuChina

As one of the underlying technologies of the blockchain,the consensus algorithm plays a vital role in ensuring security and *** a consensus algorithm for the private blockchain,Raft has better performance than the rest of the consensus algorithms,and it does not cause problems such as the concentrated hashing power,resource waste and ***,Raft can only be used in a non-byzantine environment with a small network *** order to enable Raft to be used in a large-scale network with a certain number of byzantine nodes,this paper combines Raft and credit model to propose a Raft blockchain consensus algorithm based on credit model *** the node credit evaluation phase,RBF-based support vector machine is used as the anomaly detection method,and the node credit evaluation model is *** the Trust Nodes List(TNL)mechanism is introduced to make the consensus phase in a creditable network ***,the common node is synchronized to the consensus node to update the blockchain of the entire *** show that CRaft has better throughput and lower latency than the commonly used consortium blockchain consensus algorithm PBFT(Practical Byzantine Fault Tolerance).

关键词： credit model consensus algorithm consortium blockchain P2P network

Pptcdr: Personalized Preference Transfer Cross-Domain Recommendation Based on Heterogeneous Graph Neural Networks

学校读者我要写书评

暂无评论

SSRN 2023年

作者： Zhu, Fan Xu, Xiaolong Jiangsu Key Laboratory of Big Data Security and Intelligent Processing Nanjing University of Posts and Telecommunications No. 9 Wenyuan Road Nanjing210023 China School of Computer Science Nanjing University of Posts and Telecommunications No. 9 Wenyuan Road Nanjing210023 China

Cross-domain recommendation methods aim to solve the problems of data sparsity and cold start by leveraging rich information from the source domain to enhance the recommendation accuracy of the target domain. Many methods ignore auxiliary information, which is often structured. Therefore, based on the above issue, we propose a method called PPTCDR (Personalized Preference Transfer Cross-Domain Recommendation) based on heterogeneous graph neural networks. PPTCDR introduces auxiliary information to enrich the information of the source and target domains, alleviates the problem of inaccurate embedding caused by data sparsity, and performs feature embedding of users and items separately in the source and target domains. In addition, for user preference transfer, in order to achieve personalization, a personalized bridging function is provided for each user. PPTCDR first constructs heterogeneous graphs separately in the source and target domains, and learns the latent embeddings of users and items using heterogeneous graph neural networks. Then, a meta-network is used to generate a personalized bridging function for each user. Through the personalized bridging function, the user's embedding in the source domain can be transformed to the target domain, and the transformed user embedding is used as the initial embedding of cold start users in the target domain. In this paper, we set up three cross-domain tasks on the Amazon dataset to verify and compare the performance of the experiment. PPTCDR achieves MAE and RMSE performances of 0.67 and 0.94 respectively, which are respectively improved by 23% and 18% compared to the state-of-the-art works EMCDR and PTUPCDR. © 2023, The Authors. All rights reserved.

关键词： Embeddings

Sindiff: Spoken-to-Sign Language Generation Based Transformer Diffusion Model

学校读者我要写书评

暂无评论

SSRN 2023年

作者： Liang, Wuyan Xu, Xiaolong Jiangsu Key Laboratory of Big Data Security & Intelligent Processing Nanjing University of Posts and Telecommunications No. 9 Wenyuan Road Nanjing210023 China School of Computer Science Nanjing University of Posts and Telecommunications No. 9 Wenyuan Road Nanjing210023 China

Sign language generation (SLG) is meaningful in facilitating the communication between deaf and healthy people. Recently, transformer-based diffusion models have been achieved success in human motion generation. However, it is still under-explored how to generate long-term sign language with diffusion model. Here, we propose a transformer-based diffusion framework, SinDiff, that explore all the advantages of transformer (i.e., dynamic attention and global context) for spoken-driven SLG. To meet the goal, we develop a linear biased multi-head self-attention with periodic position encoding to generalize longer-term sign language, and a linear biased cross-modal multi-head attention function to align the spoken and sign language modalities. Besides, a transformer-based spoken encoder is introduced to solve the limitations of sign data and learn the global spoken semantics. By experimental results on our testbed, we show that the word accuracy of the SinDiff for the back-translated spoken translation with ground truth spoken reach 21.02%, which has a 2.8% improvement over the 18.22% accuracy of the typical SLG method (GEN-OBT) on PHOENIX14T test set. And the generative visualization results further verify the feasibility of the proposed method. © 2023, The Authors. All rights reserved.

关键词： Semantics

Darkfight: A Detection Method of Violent Behavior in Dark Place for intelligent Monitoring System

学校读者我要写书评

暂无评论

SSRN 2022年

作者： Wang, Jun Xu, Xiaolong Nanjing University of Posts & Telecommunications Jiangsu Key Laboratory of Big Data Security & Intelligent Processing Nanjing210023 China Nanjing University of Posts & Telecommunications School of Computer Science Nanjing210023 China

With the continuous development of deep learning, the use of deep learning methods for intelligent violent behavior recognition has become an active research field in computer vision. However, there are relatively few studies on violent behavior recognition in dark light environment. In this paper, we propose the DarkFight network model for violent behavior recognition in dark light environment. In the proposed DarkFight model, Resnet50 is used as the basic network model to extract features. In order to obtain multi-scale spatiotemporal features, DarkFight adds a multi-scale attention module PSA to the network. The DarkFight model uses Zero-DCE for low-light image enhancement. In order not to lose the spatiotemporal characteristics of the low-light environment, DarkFight uses the initial dark-light data and the enhanced data as the input to the two paths of the model. In addition, this paper researches the replacement of the traditional temporal global average pooling layer of the network with the BERT module to better utilize the temporal information when fusing the two path features. We construct a violent behavior dataset VDID (Violence Detection in Dark) in a dark-light environment, and conduct experimental verification on this dataset. The verification results show that the recognition precision of the DarkFight model increases by 2.51%, the accuracy increases by 3.4%, and the F1-score improves by 0.032. © 2022, The Authors. All rights reserved.

关键词： Image enhancement

Distributed Image Retrieval Base on LSH Indexing on Spark 1st

学校读者我要写书评

暂无评论

Distributed Image Retrieval Base on LSH Indexing on Spark

1st International Conference on big data and security, ICBDS 2019

作者： Hou, Zelei Huang, Chao Wu, Jiagao Liu, Linfeng School of Computer Nanjing University of Posts and Telecommunications NanjingJiangsu210023 China Jiangsu Key Laboratory of Big Data Security and Intelligent Processing NanjingJiangsu210023 China

ISBN: (纸本)9789811575297

With the advent of the era of big data, how to process massive image, video and other multimedia data timely and accurately has become a new challenge in related fields. Aiming at the computational bottleneck and inefficiency of traditional image content retrieval system, in this paper, a distributed image retrieval framework base on Location Sensitive Hash (LSH) indexing is proposed on Spark, which combines with the distributed storage and computing characteristics of Spark big data platform. Then, distributed K-means based Bag of Visual Word (BoVW) algorithm and LSH algorithm are proposed to build LSH index vectors on Spark in parallel. The experiment shows that the retrieval time of our framework can be reduced significantly, and both retrieval recall and precision are high compared with the traditional retrieval method. © 2020, Springer Nature Singapore Pte Ltd.

关键词： Indexing (of information)