检索结果-内蒙古大学图书馆

Attention-Based Interaction Trajectory Prediction 9th

Attention-Based Interaction Trajectory Prediction

9th International Conference on Artificial Intelligence and Mobile Services, AIMS 2020, held as part of the Services Conference Federation, SCF 2020

作者： Liu, Zhe Zhang, Lizong Rao, Zhihong Liu, Guisong School of Computer Science and Engineering University of Electronic Science and Technology of China Chengdu611731 China China Electronic Technology Cyber Security Co. Ltd. Chengdu610000 China Trusted Cloud Computing and Big Data Key Laboratory of Sichuan Province Chengdu611731 China School of Computer Science Zhongshan Institute University of Electronic Science and Technology of China Zhongshan528402 China

ISBN: (纸本)9783030596040

Trajectory prediction is a hot topic in the field of computer vision and has a wide range of applications. Trajectory prediction refers to predicting the future trajectory of a target based on its past trajectory. This paper proposes a method based on graph neural network and attention mechanism, in order to update trajectory characteristics by implement global pedestrian interaction. And, a direct relationship between history and future is introduced with the attention module for reducing error propagation. The method was evaluated on several real-world crowd datasets, the results demonstrate the effectiveness of our method. © 2020, Springer Nature Switzerland AG.

关键词： Trajectories

来源：评论

学校读者我要写书评

暂无评论

OSAF_e: One-Stage Anchor Free Object Detection Method Considering Effective Area 9th

OSAF_e: One-Stage Anchor Free Object Detection Method Consid...

引用

9th International Conference on Artificial Intelligence and Mobile Services, AIMS 2020, held as part of the Services Conference Federation, SCF 2020

作者： Zhang, Yong Zhang, Lizong Rao, Zhihong Duan, Guiduo Wang, Chunyu School of Computer Science and Engineering University of Electronic Science and Technology of China Chengdu611731 China China Electronic Technology Cyber Security Co. Ltd. Chengdu610000 China Trusted Cloud Computing and Big Data Key Laboratory of Sichuan Province Chengdu611731 China School of Information and Software Engineering University of Electronic Science and Technology of China Chengdu610054 China

ISBN: (纸本)9783030596040

The task of object detection is to identify the bounding box of the object and its corresponding category in images. In this paper, we propose a new one-stage anchor free object detection algorithm OSAF_e, with the consideration of effective mapping area. A feature extraction network is used to obtain high level feature, and the true bounding box of the object in the original image is mapped to the grid of feature map, in order to perform category prediction and bounding box regression. The proposed algorithm is evaluated with the Pascal Voc dataset, and the experiments indicate that it has a better result. © 2020, Springer Nature Switzerland AG.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Relation-Based Multi-Teacher Knowledge Distillation

Relation-Based Multi-Teacher Knowledge Distillation

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Yutian Wang Yi Xu School of Information and Software Engineering University of Electronic Science and Technology of China Chengdu China Trusted Cloud Computing and Big Data Key Laboratory of Sichuan Province Chengdu China Laboratory of Intelligent Collaborative Computing University of Electronic Science and Technology of China

ISBN: (数字)9798350359312

ISBN: (纸本)9798350359329

Traditional knowledge distillation adopts one teacher model to instruct the training of a lightweight student model. To improve the performance of knowledge distillation, multi-teacher knowledge distillation utilizing multi-party knowledge to guide model learning was proposed. However, existing methods mainly use the summation-result of teacher models of a single sample to integrate multi-party knowledge directly, ignoring the relation knowledge among samples. In this paper, we propose a novel multi-teacher knowledge distillation method, which utilizes the data relation knowledge to allocate weights for teachers adaptively. To obtain the data relation knowledge, we design output-based strategy and feature-based strategy, which would help to allocate more weight for the teacher who has better learned the data relation knowledge. Extensive experiments have demonstrated the performance and efficiency of our proposed multi-teacher knowledge distillation method.

关键词： Training Adaptation models Neural networks

来源：评论

学校读者我要写书评

暂无评论

Some Thoughts on the Classification Standard of Intelligent Autonomous Level of Measurement and Control System

Some Thoughts on the Classification Standard of Intelligent ...

引用

International Conference on big data and Information Analytics (bigDIA)

作者： Guangfu Yuan Liu Yang Haoran Han Yong Li Cong Shao Siyu Zhan Gansu Photoelectric Measurement Technology Research Laboratory Gansu China Laboratory of Intelligent Collaborative Computing University of Electronic Science and Technology of China Chengdu China Trusted Cloud Computing and Big Data Key Laboratory of Sichuan Province Chengdu China

Intelligence is an important trend in the development of measurement and control systems. However, there is still a lack of unified understanding of the intelligent autonomous classification of the measurement and control system. The establishment of intelligent evaluation standards in this field has become an urgent research topic. This paper first gives an overview of the aviation measurement and control system, and then takes the classification standards of autonomy in the fields of automobile driving automation in China and the United states, the autonomy and autonomy level of typical ships in China, and the automation of civil unmanned aircraft system in China as a reference. A 5-level intelligent grading standard based on task adaptability, environmental complexity and human-computer interaction is proposed. The classification standard is practical and easy to understand. It can provide reference for the establishment of intelligent standards for measurement and control systems, improve the application speed of intelligent autonomous technology in the field of measurement and control, and broaden the application prospect of intelligent measurement and control systems.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Smoothed multi-view subspace clustering

arXiv

引用

arXiv 2021年

作者： Chen, Peng Liu, Liang Ma, Zhengrui Kang, Zhao Jangsu Automation Research Institute Jiangsu Lianyungang China University of Electronic Science and Technology of China Sichuan Chengdu China Trusted Cloud Computing and Big Data Key Laboratory of Sichuan Province Sichuan Chengdu China

In recent years, multi-view subspace clustering has achieved impressive performance due to the exploitation of complementary imformation across multiple views. However, multi-view data can be very complicated and are not easy to cluster in real-world applications. Most existing methods operate on raw data and may not obtain the optimal solution. In this work, we propose a novel multi-view clustering method named smoothed multi-view subspace clustering (SMVSC) by employing a novel technique, i.e., graph filtering, to obtain a smooth representation for each view, in which similar data points have similar feature values. Specifically, it retains the graph geometric features through applying a low-pass filter. Consequently, it produces a "clustering-friendly" representation and greatly facilitates the downstream clustering task. Extensive experiments on benchmark datasets validate the superiority of our approach. Analysis shows that graph filtering increases the separability of classes. © 2021, CC BY-NC-ND.

关键词： Low pass filters

来源：评论

学校读者我要写书评

暂无评论

Budgeted Persuasion on User Opinions via Varying Susceptibility

Budgeted Persuasion on User Opinions via Varying Susceptibil...

引用

IEEE International Conference on Performance, computing and Communications (IPCCC)

作者： Wenyi Tang Xinrui Xu Guangchun Luo Zaobo He Kaiming Zhan School of Computer Science and Engineering University of Electronic Science and Technology of China School of Information and Software Engineering University of Electronic Science and Technology of China Trusted Cloud Computing and Big Data Key Laboratory of Sichuan Province Miami University Sichuan Jiuyuan Yinhai Software Co. Ltd

ISBN: (数字)9781728198293

ISBN: (纸本)9781728198309

Nowadays, the social network becomes an indispensable part of people's daily life, meanwhile offers an unprecedentedly convenient access for purposive individuals to influence the opinions of network users. Current studies present a subtle persuasion approach that finds a number of key users meanwhile varies their susceptibility extent to impact the public opinion. Such persuasion is significantly critical for public security, as it could facilitate both the spreading and dispelling of malicious rumors. However, the major body of these studies enclose impractical assumptions, such that persuaders have an unlimited budget, or the costs of varying different users' susceptibilities are the same, thus rendering these works unsuitable for realistic scenarios. Therefore, this work originally proposes a more practical and generalized problem of persuasion, where varying the susceptibilities of different users holds different costs. The analysis of its non-convexity, non-submodularity and complexity shows that solving the proposed problem is nontrivial, thus inspiring us to provide an intuitive greedy algorithm. Furthermore, we design an accelerated algorithm based on the community property, which reduces the time consumption more than one order of magnitude. The acceleration is based on the intuition that the impact of a user within a proper community could be a good estimation of the impact in the whole network, while the computation of the former one is much more efficient. The relationship between two algorithms is fully analyzed, which shows the community-based algorithm can degenerate to the intuitive greedy algorithm under a specific setting. Finally, comprehensive evaluations on real-world datasets show the superiority of proposed algorithms on both effectiveness and efficiency.

关键词： Greedy algorithms Social networking (online) Heuristic algorithms Search problems Rendering (computer graphics) Acceleration Security

来源：评论

学校读者我要写书评

暂无评论

Pre-classification Supporting Reasoning for Document-level Relation Extraction 21

Pre-classification Supporting Reasoning for Document-level R...

引用

Proceedings of the 10th International Joint Conference on Knowledge Graphs

作者： Jiehao Zhao Guiduo Duan Tianxi Huang School of Computer Science and Engineering University of Electronic Science and Technology of China China School of Computer Science and Engineering Trusted Cloud Computing and Big Data Key Laboratory of Sichuan Province University of Electronic Science and Technology of China China Department of Fundamental Courses Chengdu Textile College China

ISBN: (纸本)9781450395656

The document-level relation extraction task aims to extract relational triples from a document consisting of multiple sentences. Most previous models focus on modeling the dependency between the entities and neglect the reasoning mechanism. Some other models construct paths implicitly between co-sentence entities to find semantic relations. However, they ignore that there are interactions between different triples, especially some triples play an import role in predicting others. In this short research paper, we propose a new two stage framework PCSR(Pre-classification Supporting Reasoning) which captures the interactions between triples and utilizes these information for reasoning. Specifically, we make a pre-classification for each entity pair in the first stage. Then we aggregate the embeddings of predicted triples to enhance entity representation and make a new classification. Since the second classification could find triples missed in the first stage, we take the result as the supplement of the prior one. Experiments on DocRED show that our method achieves an F1 score of 62.11. Compared with the previous state-of-the-art model, our model increase by 0.81 on the test set, which demonstrates the effectiveness of our reasoning mechanism.

关键词： BERT Information Extraction Document-level Relation Extraction Reasoning Mechanism

来源：评论

学校读者我要写书评

暂无评论

Relation-Guided Representation Learning

arXiv

引用

arXiv 2020年

作者： Kang, Zhao Lu, Xiao Liang, Jian Bai, Kun Xu, Zenglin School of Computer Science and Engineering University of Electronic Science and Technology of China Sichuan China Trusted Cloud Computing and Big Data Key Laboratory of Sichuan Province Cloud and Smart Industries Group Tencent Beijing China School of Computer Science and Technology Harbin Institute of Technology Shenzhen China Center for Artificial Intelligence Peng Cheng Lab Shenzhen China

Deep auto-encoders (DAEs) have achieved great success in learning data representations via the powerful representability of neural networks. But most DAEs only focus on the most dominant structures which are able to reconstruct the data from a latent space and neglect rich latent structural information. In this work, we propose a new representation learning method that explicitly models and leverages sample relations, which in turn is used as supervision to guide the representation learning. Different from previous work, our framework well preserves the relations between samples. Since the prediction of pairwise relations themselves is a fundamental problem, our model adaptively learns them from data. This provides much flexibility to encode real data manifold. The important role of relation and representation learning is evaluated on the clustering task. Extensive experiments on benchmark data sets demonstrate the superiority of our approach. By seeking to embed samples into subspace, we further show that our method can address the large-scale and out-of-sample problem. Our source code is publicly available at: https://***/nbShawnLu/RGRL. Copyright © 2020, 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

Weighted Discriminative Dictionary Learning based on Low-rank Representation

引用

Journal of Physics: Conference Series 2017年第1期787卷

作者： Heyou Chang Hao Zheng Key Laboratory of Trusted Cloud Computing and Big Data Analysis School of Information Engineering Nanjing XiaoZhuang University Nanjing China

Low-rank representation has been widely used in the field of pattern classification, especially when both training and testing images are corrupted with large noise. Dictionary plays an important role in low-rank representation. With respect to the semantic dictionary, the optimal representation matrix should be block-diagonal. However, traditional low-rank representation based dictionary learning methods cannot effectively exploit the discriminative information between data and dictionary. To address this problem, this paper proposed weighted discriminative dictionary learning based on low-rank representation, where a weighted representation regularization term is constructed. The regularization associates label information of both training samples and dictionary atoms, and encourages to generate a discriminative representation with class-wise block-diagonal structure, which can further improve the classification performance where both training and testing images are corrupted with large noise. Experimental results demonstrate advantages of the proposed method over the state-of-the-art methods.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Malicious Family Identify Combining Multi-channel Mapping Feature Image and Fine-Tuned CNN

Malicious Family Identify Combining Multi-channel Mapping Fe...

引用

IEEE International Conference on Trust, Security and Privacy in computing and Communications (TrustCom)

作者： Chenghua Tang Chen Zhou Min Hu Mengmeng Yang Baohua Qiang Guangxi Key Laboratory of Trusted Software Guilin University of Electronic Technology Guilin China Guangxi Key Laboratory of Cryptography and Information Security Guilin University of Electronic Technology Guilin China Strategic Centre for Research in Privacy-Preserving Technologies & Systems Nanyang Technological University Singapore Guangxi Cloud Computing and Big Data Collaborative Innovation Center Guilin University of Electronic Technology Guilin China

Using the features of malicious family to detect malicious code can improve the analysis efficiency and reduce the workload. Aiming at the problems of low efficiency and low accuracy of classifiers in analyzing and detecting malicious code using traditional machine learning, a new malicious family identification method 2MFI-FT is proposed. Firstly, the more comprehensive and essential features of malicious code are obtained by extracting local feature information, assembly instruction set information and visible string information. Secondly, in order to solve the problem of inconsistent images generated by feature sequences, different feature grayscale images are fused into a multi-channel mapping feature image (2MFI) based on the signature matrix. Thirdly, in view of the high cost of collecting and labeling enough data, a classification model with high recognition accuracy and strong generalization ability is realized, based on the fine-tuned Convolutional Neural Network (CNN). The regularization technique is also used to improve the robustness of the model. It is tested on Microsoft Malware Classification Challenge dataset. The experimental results show that the proposed 2MFI method for multi-channel feature image visualization has good identification effect and low time consumption when used as the input of the network model. At the same time, the 2MFI-FT method combined with fine-tuned CNN can accurately identify malicious families, and the accuracy on small training sets and large training sets can reach about 98.25% and 99.68% respectively, which provides a feasible solution for effective identification of malicious families.

关键词： Training Codes Instruction sets Semantics Feature extraction Malware Robustness

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：