检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Wang, Guiqin Zhao, Peng Zhao, Cong Yang, Shusen Cheng, Jie Leng, Luziwei Liao, Jianxing Guo, Qinghai School of Computer Science and Technology Xi’an Jiaotong University China ACS Lab Huawei Technologies China School of Mathematics and Statistics Xi’an Jiaotong University China National Engineering Laboratory for Big Data Analytics Xi’an Jiaotong University China

Weakly-supervised action localization aims to recognize and localize action instancese in untrimmed videos with only video-level labels. Most existing models rely on multiple instance learning(MIL), where the predictions of unlabeled instances are supervised by classifying labeled bags. The MIL-based methods are relatively well studied with cogent performance achieved on classification but not on localization. Generally, they locate temporal regions by the video-level classification but overlook the temporal variations of feature semantics. To address this problem, we propose a novel attention-based hierarchically-structured latent model to learn the temporal variations of feature semantics. Specifically, our model entails two components, the first is an unsupervised change-points detection module that detects change-points by learning the latent representations of video features in a temporal hierarchy based on their rates of change, and the second is an attention-based classification model that selects the change-points of the foreground as the boundaries. To evaluate the effectiveness of our model, we conduct extensive experiments on two benchmark datasets, THUMOS-14 and ActivityNetv1.3. The experiments show that our method outperforms current state-of-the-art methods, and even achieves comparable performance with fully-supervised methods. Copyright © 2023, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

On the Effectiveness of Function-Level Vulnerability Detectors for Inter-Procedural Vulnerabilities

arXiv

引用

arXiv 2024年

作者： Li, Zhen Wang, Ning Zou, Deqing Li, Yating Zhang, Ruqian Xu, Shouhuai Zhang, Chao Jin, Hai School of Cyber Science and Engineering Huazhong University of Science and Technology Wuhan China Department of Computer Science University of Colorado Colorado Springs Colorado SpringsCO United States Institute for Network Sciences and Cyberspace Tsinghua University Beijing China School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Hubei Key Laboratory of Distributed System Security Hubei Engineering Research Center on Big Data Security Cluster and Grid Computing Lab China JinYinHu Laboratory Wuhan China

Software vulnerabilities are a major cyber threat and it is important to detect them. One important approach to detecting vulnerabilities is to use deep learning while treating a program function as a whole, known as function-level vulnerability detectors. However, the limitation of this approach is not understood. In this paper, we investigate its limitation in detecting one class of vulnerabilities known as inter-procedural vulnerabilities, where the to-be-patched statements and the vulnerability-triggering statements belong to different functions. For this purpose, we create the first Inter-Procedural Vulnerability dataset (InterPVD) based on C/C++ open-source software, and we propose a tool dubbed VulTrigger for identifying vulnerability-triggering statements across functions. Experimental results show that VulTrigger can effectively identify vulnerability-triggering statements and inter-procedural vulnerabilities. Our findings include: (i) inter-procedural vulnerabilities are prevalent with an average of 2.8 inter-procedural layers;and (ii) function-level vulnerability detectors are much less effective in detecting to-be-patched functions of inter-procedural vulnerabilities than detecting their counterparts of intra-procedural vulnerabilities. Copyright © 2024, The Authors. All rights reserved.

关键词： Open source software

来源：评论

学校读者我要写书评

暂无评论

How Graph Neural Networks Learn: Lessons from Training Dynamics

arXiv

引用

arXiv 2023年

作者： Yang, Chenxiao Wu, Qitian Wipf, David Sun, Ruoyu Yan, Junchi School of Artificial Intelligence Department of Computer Science and Engineering MoE Lab of AI Shanghai Jiao Tong University China Amazon Web Services United States School of Data Science The Chinese University of Hong Kong Shenzhen China Shenzhen International Center for Industrial and Applied Mathematics Shenzhen Research Institute of Big Data China

A long-standing goal in deep learning has been to characterize the learning behavior of black-box models in a more interpretable manner. For graph neural networks (GNNs), considerable advances have been made in formalizing what functions they can represent, but whether GNNs will learn desired functions during the optimization process remains less clear. To fill this gap, we study their training dynamics in function space. In particular, we find that the gradient descent optimization of GNNs implicitly leverages the graph structure to update the learned function, as can be quantified by a phenomenon which we call kernel-graph alignment. We provide theoretical explanations for the emergence of this phenomenon in the overparameterized regime and empirically validate it on real-world GNNs. This finding offers new interpretable insights into when and why the learned GNN functions generalize, highlighting their limitations in heterophilic graphs. Practically, we propose a parameter-free algorithm that directly uses a sparse matrix (i.e. graph adjacency) to update the learned function. We demonstrate that this embarrassingly simple approach can be as effective as GNNs while being orders-of-magnitude faster. Copyright © 2023, The Authors. All rights reserved.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Enhancing Efficiency of Declarative Aggregation for Video Analytics Based on Neural Networks

Enhancing Efficiency of Declarative Aggregation for Video An...

引用

Intelligent & Innovative Practices in engineering & Management (IIPEM), International Conference on

作者： Kamalesh Thoppaen Suresh Babu Tarun Dhar Diwan G. Prasanna Lakshmi Vijayshri Khedkar Pravin B Khatkale Bhagyashree Ashok Tingare Globallogic Inc Santa Clara CA USA Atal Bihari Vajpayee University Bilaspur India School of Computer Science and Engineering Sandip University Nashik Maharashtra Department of Computer Science & Engineering Symbiosis Institute of Technology Symbiosis International (Deemed University) Pune India Sanjivani University Kopargaon Maharashtra India Department of Artificial Intelligence and Data Science D Y Patil College of Engineering Akurdi Pune

ISBN: (数字)9798350390049

ISBN: (纸本)9798350390056

The rapid expansion of video data necessitates advanced techniques for efficient processing and analysis. This paper addresses the challenges associated with declarative aggregation and limit queries in video analytics by leveraging neural networks to optimize query execution. Traditional methods often struggle with the volume and complexity of video data, leading to performance bottlenecks. Our research introduces novel algorithms that exploit neural network parallelism to enhance the efficiency of these queries. We propose a framework that integrates neural networks with custom query processing engines, focusing on spatial and temporal aggregation, as well as limit queries. Through extensive experimental studies, including real-world scenarios, we demonstrate significant improvements in query performance. Our results highlight the effectiveness of our approach in reducing processing times and increasing throughput, paving the way for more responsive and scalable video analytics systems.

关键词： Technological innovation Runtime Visual analytics Query processing Neural networks Machine learning Parallel processing Streaming media Throughput Real-time systems

来源：评论

学校读者我要写书评

暂无评论

On the Effectiveness of Function-Level Vulnerability Detectors for Inter-Procedural Vulnerabilities

On the Effectiveness of Function-Level Vulnerability Detecto...

引用

International Conference on Software engineering (ICSE)

作者： Zhen Li Ning Wang Deqing Zou Yating Li Ruqian Zhang Shouhuai Xu Chao Zhang Hai Jin National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Hubei Key Laboratory of Distributed System Security Hubei Engineering Research Center on Big Data Security Cluster and Grid Computing Lab School of Cyber Science and Engineering Huazhong University of Science and Technology Wuhan China Jin YinHu Laboratory Wuhan China Department of Computer Science University of Colorado Colorado Springs Colorado Springs Colorado USA Institute for Network Sciences and Cyberspace Tsinghua University Beijing China School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China

ISBN: (数字)9798400702174

ISBN: (纸本)9798350382143

Software vulnerabilities are a major cyber threat and it is important to detect them. One important approach to detecting vulnerabilities is to use deep learning while treating a program function as a whole, known as function-level vulnerability detectors. However, the limitation of this approach is not understood. In this paper, we investigate its limitation in detecting one class of vulnerabilities known as inter-procedural vulnerabilities, where the to-be-patched statements and the vulnerability-triggering statements belong to different functions. For this purpose, we create the first Inter -Procedural Vulnerability dataset (InterPVD) based on C/C++ open-source software, and we propose a tool dubbed VulTrigger for identifying vulnerability-triggering statements across functions. Experimental results show that VulTrigger can effectively identify vulnerability-triggering statements and inter-procedural vulnerabilities. Our findings include: (i) inter-procedural vulnerabilities are prevalent with an average of 2.8 inter-procedural layers; and (ii) function-level vulner-ability detectors are much less effective in detecting to-be-patched functions of inter-procedural vulnerabilities than detecting their counterparts of intra-procedural vulnerabilities.

关键词： Deep learning Detectors Open source software Software engineering

来源：评论

学校读者我要写书评

暂无评论

Target-Oriented WiFi Sensing for Respiratory Healthcare: from Indiscriminate Perception to In-Area Sensing

引用

IEEE Network 2024年

作者： Wang, Meng Huang, Jinyang Zhang, Xiang Liu, Zhi Li, Meng Zhao, Peng Yan, Huan Sun, Xiao Dong, Mianxiong Hefei University of Technology Anhui Province Key Laboratory of Affective Computing and Advanced Intelligence Machine School of Computer and Information Hefei230601 China University of Science and Technology of China CAS Key Laboratory of Electromagnetic Space Information Hefei230026 China The University of Electro-Communications Department of Computer and Network Engineering Tokyo1828585 Japan Guizhou Normal University School of Big Data and Computer Science Guiyang550025 China Muroran Institute of Technology Department of Sciences and Informatics 0508585 Japan

Driven by the vision of integrated sensing and communication (ISAC) toward 6G technology, the WiFi-based respiration sensing approach has emerged as a highly competitive candidate for advanced healthcare services. Nevertheless, the indiscriminate perception of all the moving objects within the sensing area raises challenges for system stability and robustness for real-world deployment, especially for the susceptibility to motion interference from other people. Meanwhile, thanks to the emerging in-area wireless sensing technologies, i.e., beamforming, the dynamic environment tolerance of these technologies gives crucial opportunities for target-oriented WiFi-based healthcare services. This article discusses the compelling intersection of in-area WiFi sensing and the future of target-oriented intelligent respiratory healthcare network services. We first present a novel WiFi-based healthcare-assisting framework tailored for inter-user motion interference-tolerable (IMIT) sensing. The detailed components of low-level signal preprocessing, target-direction signal extraction, and healthcare-related services are discussed. By utilizing CSI beamforming technology, which adjusts a directional beam toward the desired direction and adaptively places beam nulls in noisy directions, we can mitigate motion interference and achieve target-oriented healthcare sensing. Finally, we discuss open challenges and potential solutions for target-oriented WiFi healthcare sensing. © 1986-2012 IEEE.

关键词： Wireless local area networks (WLAN)

来源：评论

学校读者我要写书评

暂无评论

Hybrid Movie Recommendation System using Neural Collaborative Filtering with Gradient Decent and Association Rule Mining

Hybrid Movie Recommendation System using Neural Collaborativ...

引用

Software, Systems and Information Technology (SSITCON), International Conference on

作者： J Cindhamani Rajkumar Pogaku M. P Sahana Myasar M. Adnan M. Suganya Department of Artificial Intelligence and Data Science Vel Tech Multi Tech Dr. Rangrajan Dr. Sakunthala Engineering College Avadi Chennai India School of Computer Science and Engineering Malla Reddy Engineering College for Women (UGC-Autonomous) Maisammaguda Hyderabad India Department of Computer Science and Engineering Dayananda Sagar College of Engineering Bengaluru India Department of Computers Techniques Engineering College of Technical Engineering The Islamic University Najaf Iraq Department of Electronics and Communication Engineering New Prince Shri Bhavani College of Engineering and Technology Chennai India

ISBN: (数字)9798350352931

ISBN: (纸本)9798350352948

In recent days, most of the people are entertained by watching movies and shows, so developing a recommendation system help people across the world in saving the time to search for high rated movies. Usually, movies are recommended based on the user interests in the different movie genres such as horror, drama, romantic and comedy. Previous researchers suggested various methods for movie recommendation using Content Based Filtering (CBF), Collaborative Filtering (CF) which failed to predict the non-linear relationship recommendation due to the availability of many genres. To overcome this challenge, we proposed hybrid model namely Neural Collaborative Filtering (NCF) with gradient decent and Association Rule Mining (ARM). Initially, MovieLens dataset is employed and then data cleaning data organization is incorporated in pre-processing. Further, the features are extracted through the Term Frequency-Inverse Document Frequency (TF-IDF) and Latent Semantic Analysis (LSA) to convert the text data into numerical data and increase the performance respectively. Finally, the developed model was measured using Root mean square error (RMSE), precision, recall, and F 1 score. Hence, the proposed hybrid recommendation model provides higher accuracy value when compared to Knowledge-aware Attentional Neural Network (KANN).

关键词： Collaborative filtering Semantics Neural networks Organizations Motion pictures Feature extraction Frequency conversion Numerical models Association rule learning Recommender systems

来源：评论

学校读者我要写书评

暂无评论

Covert Communications with Enhanced Physical Layer Security in RIS-Assisted Cooperative Networks

引用

IEEE Transactions on Wireless Communications 2025年

作者： Li, Xingwang Liu, Musen Dang, Shuping Luong, Nguyen Cong Yuen, Chau Nallanathan, Arumugam Niyato, Dusit Henan Polytechnic University School of Physics and Electronic Information Engineering Jiaozuo454003 China Southeast University National Mobile Communications Research Laboratory Nanjing210096 China Jiaozuo Key Laboratory of Crow-Sensing Network Jiaozuo454003 China University of Bristol School of Electrical Electronic and Mechanical Engineering BristolBS8 1UB United Kingdom Phenikaa University Faculty of Computer Science Hanoi12116 Viet Nam Nanyang Technological University Singapore School of Electrical and Electronics Engineering 639798 Singapore Queen Mary University of London School of Electronic Engineering and Computer Science London United Kingdom Kyung Hee University Department of Electronic Engineering Gyeonggi-do Yongin-si17104 Korea Republic of Nanyang Technological University Singapore College of Computing and Data Science 639798 Singapore

Reconfigurable intelligent surface (RIS) and ambient backscatter communication (AmBC) technologies are recognized for their programmability and high energy efficiency respectively, which will be the key parts of the future sixth generation (6G) mobile communication technology. The combination of the two technologies can improve communication security by reducing the probability of detection and decoding through enhanced transmission and backscatter transmission in different communication slots. In this paper, a dual-function RIS that supports cooperative relaying for covert communications is proposed. It operates in different communication slots (enhanced transmission slot and backscatter slot), but the performance is affected by phase errors due to function switching. A source covertly communicates with an intended destination via the help of RIS and cooperative relay. There is an illegal monitor aims to detect and eavesdrop the covert message. For this system, the outage probability (OP), intercept probability (IP), and detection error probability (DEP) in different communication slots are derived to examine the system reliability and security. Moreover, the system security probability (SSP) is proposed, and a block coordinated ascent (BCA)-based iterative algorithm is used to jointly optimize the power allocation coefficients to maximize the SSP. Simulation results show that increasing the number of elements can improve the security performance and mitigate the negative impact of RIS phase errors. © 2002-2012 IEEE.

关键词： Mobile telecommunication systems

来源：评论

学校读者我要写书评

暂无评论

PDEGAN: A Panoramic Style Transfer Based on Generative Adversarial Networks

Journal of Network Intelligence

引用

Journal of Network Intelligence 2024年第4期9卷 2112-2121页

作者： Wang, Qinghua Long, Xinling Huang, Jingwei Chen, Yang Yang, Lirong Zhang, Fuquan College of Computer and Big Data Minjiang University Fuzhou350108 China Fuzhou Technology Innovation Center of intelligent Manufacturing information System Minjiang University Fuzhou350108 China Fujian Province University Fuzhou350300 China College of Computer and Control Engineering Minjiang University Fuzhou350108 China College of Computer and Big Data Fuzhou University Fuzhou China School of Mechanical and Automotive Engineering Fujian University of Technology Fuzhou China College of foreign languages Dalian Jiaotong University Dalian116028 China Digital Media Art Key Laboratory of Sichuan Province Sichuan Conservatory of Music Chengdu610021 China

Panoramic image plays an extremely important role in the application of 3D, but in some special scenarios, the depth information is also needed as an auxiliary, but to obtain the panoramic depth information needs a high cost, it is extremely difficult to obtain. Therefore, this paper proposes a panoramic depth estimation method called PDEGAN (Panoramic Depth Estimation Generative Adversarial Network). This method first enhances the style migration, transfers the input panoramic image to a style closer to the panoramic depth image domain, and then combines the conversion results with the original panoramic image as the input data for further depth estimation training. By using the style closer to the depth estimation, the texture, color and other characteristics of the image can be adjusted, so as to better highlight the depth information of the object in the image, so that the generator network is more focused on performing the image translation task, reduce the impact of style differences on the image translation task, and improve the generalization ability of the model. At the same time, the model will also use multi-scale feature fusion, combined with the attention mechanism, to better capture the depth information of different scales, viewing angles and channels. Not only that, but also by introducing depth gradient loss, surface normal loss to improve the accuracy and stability of depth estimation. To this end, we also made a set of related datasets for the related experiments of PDEGAN, and named it PDE Images. Finally, it is shown that the method proposed in this paper has more accurate estimation results than the traditional Pix2Pix image conversion model. © 2024, Taiwan Ubiquitous Information CO LTD. All rights reserved.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

SoAy: A Solution-based LLM API-using Methodology for Academic Information Seeking 25

SoAy: A Solution-based LLM API-using Methodology for Academi...

引用

Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and data Mining V.1

作者： Yuanchun Wang Jifan Yu Zijun Yao Jing Zhang Yuyang Xie Shangqing Tu Yiyang Fu Youhe Feng Jinkai Zhang Jingyao Zhang Bowen Huang Yuanyao Li Huihui Yuan Lei Hou Juanzi Li Jie Tang School of Information Renmin University of China Beijing China & Key Laboratory of Data Engineering and Knowledge Engineering MOE Beijing China Institute of Education Tsinghua University Beijing China Department of Computer Science and Technology Tsinghua University Beijing China School of Information Renmin University of China Beijing China & Engineering Research Center of Database and Business Intelligence MOE Beijing China School of Information Renmin University of China Beijing China Zhipu AI Beijing China

ISBN: (纸本)9798400712456

Applying large language models (LLMs) to academic API usage shows promise in reducing researchers' efforts to seek academic information. However, current LLM methods for using APIs struggle with the complex API coupling commonly encountered in academic queries. To address this, we introduce SoAy, a solution-based LLM methodology for academic information seeking. SoAy enables LLMs to generate code for invoking APIs, guided by a pre-constructed API calling sequence referred to as a solution. This solution simplifies the model's understanding of complex API relationships, while the generated code enhances reasoning efficiency. LLMs are aligned with this solution-oriented, code-based reasoning method by automatically enumerating valid API coupling sequences and transforming them into queries and executable *** evaluate SoAy, we introduce SoAyBench, an evaluation benchmark accompanied by SoAyEval, built upon a cloned environment of APIs from AMiner. Experimental results demonstrate a 34.58-75.99% performance improvement compared to state-of-the-art LLM API-based baselines. All datasets, codes, tuned models, and deployed online services are publicly accessible at https://***/RUCKBReasoning/SoAy.

关键词： academic information seeking

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：