检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Hu, Yuxue Li, Junsong Chen, Meixuan Su, Dongyu Wang, Tongguan Sha, Ying College of Informatics Huazhong Agricultural University Wuhan China Key Laboratory of Smart Farming for Agricultural Animals Wuhan China Hubei Engineering Technology Research Center of Agricultural Big Data Wuhan China Engineering Research Center of Intelligent Technology for Agriculture Ministry of Education China

Euphemism identification deciphers the true meaning of euphemisms, such as linking "weed" (euphemism) to "marijuana" (target keyword) in illicit texts, aiding content moderation and combating underground markets. While existing methods are primarily text-based, the rise of social media highlights the need for multimodal analysis, incorporating text, images, and audio. However, the lack of multimodal datasets for euphemisms limits further research. To address this, we regard euphemisms and their corresponding target keywords as keywords and first introduce a keyword-oriented multimodal corpus of euphemisms (KOM-Euph), involving three datasets (Drug, Weapon, and Sexuality), including text, images, and speech. We further propose a keyword-oriented multimodal euphemism identification method (KOM-EI), which uses cross-modal feature alignment and dynamic fusion modules to explicitly utilize the visual and audio features of the keywords for efficient euphemism identification. Extensive experiments demonstrate that KOM-EI outperforms state-of-the-art models and large language models, and show the importance of our multimodal datasets 1 Copyright © 2025, The Authors. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A modeling framework for assessing the future changes in the occurrence of extreme rain-induced landslides

引用

Gondwana Research 2025年 143卷 52-63页

作者： Zhang, Xinlong Lin, Qigen Lok, Manhoi Huang, Taosheng Yu, Xuan Wang, Weiping Shen, Ping State Key Laboratory of Internet of Things for Smart City and Department of Civil and Environmental Engineering University of Macau Macao SAR China Institute for Disaster Risk Management School of Geographical Sciences Nanjing University of Information Science & Technology Nanjing China Department of Civil and Environmental Engineering University of Macau Macao SAR China Center for Water Resources and Environment School of Civil Engineering Sun Yat-sen University Guangzhou China School of National Safety and Emergency Management Beijing Normal University Zhuhai China Joint International Research Laboratory of Catastrophe Simulation and Systemic Risk Governance Beijing Normal University Zhuhai China

Landslides triggered by extreme rainfall pose a serious threat to lives and livelihoods under climate change. However, the influence of climate change on the frequency and intensity of extreme rainfall and the resulting landslide hazard remains insufficiently investigated. Here we develop a novel framework to assess how future changes in extreme rainfall affect landslide occurrence, using a comprehensive dataset of landslide events in Guangdong province, China. We apply the XGBoost machine learning algorithm with a correction by feature interaction constraints to model landslide hazards based on various predisposing and triggering factors. We use a new extreme rainfall correction method to account for the global climate model-projected changes in rainfall patterns under different emission scenarios, based on which we perform a multi-model ensemble analysis to reduce the uncertainty of climate projections. We find that regions with high landslide probability are strongly associated with intense rainfall events, especially in steep and erodible river valleys. Our projections show that under a high-emission scenario, extreme rainfall events and landslide hazards will increase substantially in Guangdong, highlighting the urgency of adaptation of climate change and risk management. Our framework provides a valuable tool for understanding the interactions between climate change, extreme rainfall, and landslide occurrence, and for developing effective mitigation strategies. © 2025 International Association for Gondwana Research

关键词： Climate change Extreme rainfall Machine Learning Rain-induced landslides

来源：评论

学校读者我要写书评

暂无评论

Energy Minimization Oriented Hybrid Semantic Data Transmission in Air-Ocean Integrated Networks: A Resource Allocation Design

引用

IEEE Transactions on Mobile Computing 2025年

作者： Dai, Minghui Wang, Tianshun Chang, Shan Su, Zhou Wu, Yuan Donghua University School of Computer Science and Technology Shanghai201620 China Nanjing University of Posts and Telecommunications School of Communication and Information Engineering Nanjing China Xi'an Jiaotong University School of Cyber Science and Engineering Xi'an710049 China University of Macau State Key Laboratory of Internet of Things for Smart City China

With the development of new generation communication technologies, the future maritime information networks pave the way to promote the exploration of ocean resources. Moreover, the underwater data center (UDC) is considered to be a significant data storage and computing unit in future maritime networks for providing ocean services. However, the current deployment of UDC faces the critical issues, i.e., the long-distance underwater transmission is unreliable and the energy consumption and resources of underwater transmission are overloaded. To address the two critical issues of unreliable data transmission and high resource overheads, in this paper, we present a hybrid semantic data transmission architecture in air-ocean integrated networks, which can perceive the sea surface data accurately and transmit it to the UDC for processing. Specifically, in surface layer, unmanned aerial vehicles (UAVs) perceive ocean environment and send data to the buoy via non-orthogonal multiple-access (NOMA) transmission to improve the channel utilization. In underwater layer, the buoy sends the collected data to UDC via semantic transmission, while the semantic fidelity metric is utilized to improve the transmission efficiency. A resource allocation problem for energy minimization is formulated to jointly optimize the semantic scaling factor, the NOMA decoding order, the communication and computing resource allocations. We exploit a decomposition approach to transform the problem into two sub-problems, where the optimal resource allocations are obtained by proposing efficient algorithms. Finally, we provide simulations to verify the effectiveness and efficiency of our proposed scheme. The results demonstrate that our proposal has the advantages of lower energy consumption compared to several baseline schemes. © 2002-2012 IEEE.

关键词： Unmanned aerial vehicles (UAV)

来源：评论

学校读者我要写书评

暂无评论

CoT-Drive: Efficient Motion Forecasting for Autonomous Driving with LLMs and Chain-of-Thought Prompting

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2025年

作者： Liao, Haicheng Kong, Hanlin Wang, Bonan Wang, Chengyue Ye, Wang He, Zhengbing Xu, Chengzhong Li, Zhenning University of Macau State Key Laboratory of Internet of Things for Smart City Macau China University of Macau Department of Computer and Information Science Macau China Massachusetts Institute of Technology Senseable City Lab CambridgeMA United States

Accurate motion forecasting is crucial for safe autonomous driving (AD). This study proposes CoT-Drive, a novel approach that enhances motion forecasting by leveraging large language models (LLMs) and a chain-of-thought (CoT) prompting method. We introduce a teacher-student knowledge distillation strategy to effectively transfer LLMs’ advanced scene understanding capabilities to lightweight language models (LMs), ensuring that CoT-Drive operates in real-time on edge devices while maintaining comprehensive scene understanding and generalization capabilities. By leveraging CoT prompting techniques for LLMs without additional training, CoT-Drive generates semantic annotations that significantly improve the understanding of complex traffic environments, thereby boosting the accuracy and robustness of predictions. Additionally, we present two new scene description datasets, Highway-Text and Urban-Text, designed for fine-tuning lightweight LMs to generate context-specific semantic annotations. Comprehensive evaluations of five real-world datasets demonstrate that CoT-Drive outperforms existing models, highlighting its effectiveness and efficiency in handling complex traffic scenarios. Overall, this study is the first to consider the practical application of LLMs in this field. It pioneers the training and use of a lightweight LLM surrogate for motion forecasting, setting a new benchmark and showcasing the potential of integrating LLMs into AD systems. © 2020 IEEE.

关键词： Autonomous vehicles

来源：评论

学校读者我要写书评

暂无评论

Anomize: Better Open Vocabulary Video Anomaly Detection

arXiv

引用

arXiv 2025年

作者： Li, Fei Liu, Wenxuan Chen, Jingjing Zhang, Ruixu Wang, Yuran Zhong, Xian Wang, Zheng National Engineering Research Center for Multimedia Software School of Computer Science Wuhan University China Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University China State Key Laboratory for Multimedia Information Processing School of Computer Science Peking University China Hubei Key Laboratory of Transportation Internet of Things Wuhan University of Technology China

Open Vocabulary Video Anomaly Detection (OVVAD) seeks to detect and classify both base and novel anomalies. However, existing methods face two specific challenges related to novel anomalies. The first challenge is detection ambiguity, where the model struggles to assign accurate anomaly scores to unfamiliar anomalies. The second challenge is categorization confusion, where novel anomalies are often misclassified as visually similar base instances. To address these challenges, we explore supplementary information from multiple sources to mitigate detection ambiguity by leveraging multiple levels of visual data alongside matching textual information. Furthermore, we propose incorporating label relations to guide the encoding of new labels, thereby improving alignment between novel videos and their corresponding labels, which helps reduce categorization confusion. The resulting Anomize framework effectively tackles these issues, achieving superior performance on UCF-CRIME and XD-VIOLENCE datasets, demonstrating its effectiveness in OVVAD. Copyright © 2025, The Authors. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Multi-Hop Attention Diffusion Graph Neural Networks For Multimodal Fake News Detection

Multi-Hop Attention Diffusion Graph Neural Networks For Mult...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Zhongqiang Huang Dongli Lu Ying Sha Key Laboratory of Smart Farming for Agricultural Animals Wuhan China Engineering Research Center of Intelligent Technology for Agriculture Ministry of Education Wuhan China Hubei Engineering Technology Research Center of Agricultural Big Data Wuhan China College of Informatics Huazhong Agricultural University Wuhan China

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Fake news often spreads rapidly on social media in various forms, exposing users to a large amount of misinformation and disinformation. Recently, promising results have been achieved in fake news detection. However, most of the approaches only consider one-hop neighborhood information in cross-modal mapping relationships or focus mainly on pairwise interactions of multimodal fake news. However, the fine-grained feature mapping of multimodal fake news may go beyond binary relations. To address these issues, we propose a Multi-Hop Attention Diffusion Graph Neural Networks in this work to capture and complement the multivariate mapping relationships between fine-grained features, while extracting cross-modal higher-order alignment representations for multimodal fake news detection via cross-modal contrastive learning and modal matching. Extensive experiments on three public datasets demonstrate the effectiveness of the proposed approach in the multimodal fake news detection task.

关键词： Visualization Social networking (online) Fuses Spread spectrum communication Contrastive learning Signal processing Feature extraction Graph neural networks Fake news Speech processing

来源：评论

学校读者我要写书评

暂无评论

Model-Driven Channel Estimation Network for Orthogonal Time-Frequency Space Systems

引用

IEEE Transactions on Vehicular technology 2025年

作者： Wang, Qian Chen, Xingke Tao, Qin Qian, Li Ping Kam, Pooi-Yuen Wu, Yuan Zhejiang University of Technology Institute of Cyberspace Security Hangzhou310023 China Hangzhou Normal University School of Information Science and Technology Hangzhou311121 China School of Science and Engineering Shenzhen518172 China Center for Advanced Microelectronic Devices Suzhou215123 China University of Macau State Key Laboratory of Internet of Things for Smart City Department of Computer and Information Science China

Orthogonal time-frequency space (OTFS) modulation has garnered significant attention in wireless communications, particularly in high-mobility scenarios, due to its exceptional resilience to Doppler shifts. While the delay- Doppler domain channel estimation is mostly considered for reliable communication, it is often hindered by fractional Doppler shifts in complex channel conditions. To overcome this challenge, this paper proposes a model-driven channel estimation network (CENet) for accurate channel estimation in the delay-time (DT) domain, with the initial least-squares (LS) solutions as prior information for training. Specifically, a LS-driven CENet with parallel structure is proposed to efficiently handle the high-dimensional DT domain channel matrix, which enables the row- wise training and estimation with fast convergence. Simulation results demonstrate that our proposed CENet significantly outperforms the conventional LS and deep neural network-based methods. © 1967-2012 IEEE.

关键词： Least squares approximations

来源：评论

学校读者我要写书评

暂无评论

New algorithm for variable-rate linear broadcast network coding

引用

Journal of Central South University 2011年第4期18卷 1193-1199页

作者：夏寅张惕远黄佳庆 Department of Electronics and Information Engineering Huazhong University of Science and Technology Hubei Provincial Key Laboratory of Smart Internet Huazhong University of Science and Technology

To adjust the variance of source rate in linear broadcast networks, global encoding kernels should have corresponding dimensions to instruct the decoding process. The algorithm of constructing such global encoding kernels is to adjust heterogeneous network to possible link failures. Linear algebra, graph theory and group theory are applied to construct one series of global encoding kernels which are applicable to all source rates. The effectiveness and existence of such global encoding kernels are proved. Based on 2 information flow, the algorithm of construction is explicitly given within polynomial time O（｜E｜｜T｜.ω^2max）, and the memory complexity of algorithm is O（｜E｜）. Both time and memory complexity of this algorithm proposed can be O（ωmax） less than those of algorithms in related works.

关键词： network coding variable-rate linear broadcast heterogeneous network code construction algorithm

来源：评论

学校读者我要写书评

暂无评论

Unveiling Security Weaknesses in Autonomous Driving Systems: An In-Depth Empirical Study

arXiv

引用

arXiv 2025年

作者： Cheng, Wenyuan Li, Zengyang Liang, Peng Mo, Ran Liu, Hui School of Computer Science Hubei Provincial Key Laboratory of Artificial Intelligence and Smart Learning Central China Normal University Wuhan China School of Computer Science Wuhan University Wuhan China School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan China

Context: The advent of Autonomous Driving Systems (ADS) has marked a significant shift towards intelligent transportation, with implications for public safety and traffic efficiency. While these systems integrate a variety of technologies and offer numerous benefits, their security is paramount, as vulnerabilities can have severe consequences for safety and trust. Objective: This study aims to systematically investigate potential security weaknesses in the codebases of prominent open-source ADS projects using CodeQL, a static code analysis tool. The goal is to identify common vulnerabilities, their distribution and persistence across versions to enhance the security of ADS. Methods: We selected three representative open-source ADS projects, Autoware, AirSim, and Apollo, based on their high GitHub star counts and Level 4 autonomous driving capabilities. Using CodeQL, we analyzed multiple versions of these projects to identify vulnerabilities, focusing on CWE categories such as CWE-190 (Integer Overflow or Wraparound) and CWE-20 (Improper Input Validation). We also tracked the lifecycle of these vulnerabilities across software versions. This approach allows us to systematically analyze vulnerabilities in projects, which has not been extensively explored in previous ADS research. Results: Our analysis revealed that specific CWE categories, particularly CWE-190 (59.6%) and CWE-20 (16.1%), were prevalent across the selected ADS projects. These vulnerabilities often persisted for over six months, spanning multiple version iterations. The empirical assessment showed a direct link between the severity of these vulnerabilities and their tangible effects on ADS performance. Conclusions: These security issues among ADS still remain to be resolved. Our findings highlight the need for integrating static code analysis into ADS development to detect and mitigate common vulnerabilities. Meanwhile, proactive protection strategies, such as regular update of third-party libraries, are

关键词： Open source software

来源：评论

学校读者我要写书评

暂无评论

Semi-Passive IRS Enabled Sensing with Group Movable Sensors

引用

IEEE Communications Letters 2025年

作者： Peng, Qiaoyan Wu, Qingqing Chen, Wen Chen, Guangji Gao, Ying Xu, Lexi Ma, Shaodan Shanghai Jiao Tong University Department of Electronic Engineering Shanghai200240 China University of Macau State Key Laboratory of Internet of Things for Smart City Macao999078 China Nanjing University of Science and Technology School of Electronic and Optical Engineering Nanjing210094 China Southeast University National Mobile Communications Research Laboratory China China United Network Communications Corporation Research Institute Beijing100048 China

The performance of the sensing system is limited by the signal attenuation and the number of receiving components. In this letter, we investigate the sensor position selection in a semi-passive intelligent reflecting surface (IRS) enabled non-line-of-sight (NLoS) sensing system. The IRS consists of passive elements and active sensors, where the sensors can receive and process the echo signal for direction-of-arrival (DoA) estimation. Motivated by the movable antenna array and fluid antenna system, we consider the case where the sensors are integrated into a group for movement and derive the corresponding Cramer-Rao bound (CRB). Then, the optimal solution for the positions of the movable sensors (MSs) to the CRB minimization problem is derived in closed form. Moreover, we characterize the relationship between the CRB and system parameters. Theoretical analysis and numerical results are provided to demonstrate the superiority of the proposed MS scheme over the fixed-position (FP) scheme. © 1997-2012 IEEE.

关键词： Cramer-Rao bounds

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：