检索结果-内蒙古大学图书馆

IEEE Transactions on Artificial Intelligence 2025年

作者： Huang, Baojin Ma, Jiaqi Wang, Guangcheng Wang, Hui Huazhong Agricultural University College of Informatics Wuhan430070 China Wuhan University National Engineering Research Center for Multimedia Software School of Computer Wuhan430072 China Wuhan University School of Computer Science Wuhan430072 China Nantong University School of Transportation and Civil Engineering Nantong226019 China Sichuan University College of Computer Science Chengdu610065 China

Face swapping aims to replace a source face with a target face, generating a fake face that is indistinguishable from the real one to the human eye. Existing face recognition methods usually discriminate the fake face as the target face identity, which happens to be misguided. To address this embarrassment, we pioneer a new task called "fake face recognition", which seeks to discover the identity of the source face based on the fake face. Besides, we design a hybrid and dynamic learning strategy for fake face recognition. Specifically, we hybridize the existing real face recognition dataset with the fake face dataset. Based on the popular margin-based face recognition approach, we achieve dynamic learning by adjusting the margin for the fake face samples. The deep network is guided to first focus on real samples and then explore the identity implicit commonalities between real and fake samples. To verify the performance of the fake face recognition model, we further organize the existing fake face datasets into face pairs. Extensive experiments on the fake face datasets show that our proposed hybrid and dynamic learning strategy achieves superior average accuracy (98.46%) compared to benchmark studies. © 2020 IEEE.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

SINdex: Semantic INconsistency Index for Hallucination Detection in LLMs

arXiv

引用

arXiv 2025年

作者： Abdaljalil, Samir Kurban, Hasan Sharma, Parichit Serpedin, Erchin Atat, Rachad Electrical and Computer Engineering Texas A&M University College StationTX United States College of Science and Engineering Hamad Bin Khalifa University Doha Qatar Department of Computer Science Luddy School of Informatics BloomingtonIN United States Department of Computer Science and Mathematics Lebanese American University Beirut Lebanon

Large language models (LLMs) are increasingly deployed across diverse domains, yet they are prone to generating factually incorrect outputs—commonly known as "hallucinations." Among existing mitigation strategies, uncertainty-based methods are particularly attractive due to their ease of implementation, independence from external data, and compatibility with standard LLMs. In this work, we introduce a novel and scalable uncertainty-based semantic clustering framework for automated hallucination detection. Our approach leverages sentence embeddings and hierarchical clustering alongside a newly proposed inconsistency measure, SINdex, to yield more homogeneous clusters and more accurate detection of hallucination phenomena across various LLMs. Evaluations on prominent open- and closed-book QA datasets demonstrate that our method achieves AUROC improvements of up to 9.3% over state-of-the-art techniques. Extensive ablation studies further validate the effectiveness of each component in our framework. © 2025, CC BY.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Relip: Reliable in-Band Parallel Communication for Magnetic MIMO Wireless Power Transfer System

引用

IEEE Transactions on Mobile Computing 2025年

作者： Wang, Xinyu Zhou, Wangqiu Zhou, Hao Jiang, Shenyao Liu, Zhi Wang, Xiaoyan Ji, Yusheng Song, Qi City University of Hong Kong Department of Computer Science Hong Kong University of Science and Technology of China School of Computer Science and Technology Hefei230027 China The University of Electro-Communications Graduate School of Informatics and Engineering Tokyo182-8585 Japan Ibaraki University Graduate School of Science and Engineering Ibaraki310-8512 Japan National Institute of Informatics Information Systems Architecture Science Research Division Tokyo101-0003 Japan

In magnetic resonant coupling (MRC) based wireless power transfer (WPT) systems, receiver (RX) feedback communication is promising to enhance the capability and efficiency of the system. Although some studies have explored in-band implementations with low overhead costs, it has not been comprehensively investigated. In this paper, we propose Relip, a Reliable layer-level in-band parallel feedback communication mechanism for MIMO MRC-WPT systems, which addresses the impact of RX-RX couplings (i.e., non-negligible interference from strong couplings and positive effects of relay phenomenon), and provides a theoretical analysis of communication reliability. Technically, we first devise an On-Off based two-phase modulation mechanism to achieve RX identification and dependency detection under relay phenomenon. Then, we utilize observed channel decomposability to collect group-level power transfer channel conditions for eliminating the interference caused by strong RX-RX couplings. Furthermore, we perform RX selection to optimize the trade-off between communication reliability and time overhead. We design and implement the Relip prototype and conduct extensive experiments. The results validate the effectiveness of our mechanism, i.e., Relip can provide ≥99% average decoding accuracy for concurrent feedback communication of 14 devices, achieving an 18.31% improvement compared to the state-of-the-art solution. © 2002-2012 IEEE.

关键词： Inductive power transmission

来源：评论

学校读者我要写书评

暂无评论

Neural Dynamics for Constrained Bi-Objective Quadratic Programming with Applications to Scientific Computing

引用

Tsinghua Science and Technology 2025年第5期30卷 2014-2028页

作者： Xinwei Cao Xujin Pu Cheng Hua Bolin Liao Ameer Hamza Khan School of Business Jiangnan University Wuxi China College of Computer Science and Engineering Jishou University Jishou China Department of Land Surveying and Geo-Informatics (LSGI) The Hong Kong Polytechnic University Hong Kong China

Neural dynamics is a powerful tool to solve online optimization problems and has been used in many applications. However, some problems cannot be modelled as a single objective optimization and neural dynamics method does not apply. This paper proposes the first neural dynamics model to solve bi-objective constrained quadratic program, which opens the avenue to extend the power of neural dynamics to multi-objective optimization. We rigorously prove that the designed neural dynamics is globally convergent and it converges to the optimal solution of the bi-objective optimization in Pareto sense. Illustrative examples on bi-objective geometric optimization are used to verify the correctness of the proposed method. The developed model is also tested in scientific computing with data from real industrial data with demonstrated superior to rival schemes.

关键词： Scientific computing Computational modeling Linear programming Time measurement Data models Dynamic programming Quadratic programming Optimization Portfolios

来源：评论

学校读者我要写书评

暂无评论

Incorporating multilevel dynamic game and robust multi-objective optimization for cross-border river governance: A Lancang–Mekong case study

引用

Journal of Cleaner Production 2025年 513卷

作者： Yang, Ruijia Feng, Jingchun Tang, Jiansong Saga, Ryosuke Business School Hohai University Nanjing211100 China College of Computer Science and Software Engineering Hohai University Nanjing211100 China of Informatics Osaka Metropolitan University Osaka559-8531 Japan

Transboundary river governance faces increasingly complex economic, ecological, and political challenges, yet existing single-round approaches often fail to capture the dynamic negotiations and extreme uncertainty inherent in multi-country resource management. This paper proposes an integrated decision-making framework that couples a multilevel dynamic game with robust multi-objective optimization in the Lancang–Mekong River Basin, aiming to generate politically feasible and stable water-allocation strategies even under severe drought or market shocks. First, an initial robust optimization identifies a Pareto set balancing economic returns, equitable distribution, and ecological flow requirements. These solutions feed into a multi-round dynamic game model, which simulates strategic adjustments and compensation mechanisms among upstream and downstream countries. A subsequent second-stage optimization incorporates negotiation feedback to further improve both political acceptability and resilience. Results indicate that compensation incentives can effectively induce upstream stakeholders to increase critical dry-season releases by 5–10 %, thereby raising downstream ecological flow satisfaction by 10–15 percentage points and boosting agricultural production by up to 8–10 % under drought conditions. Sensitivity analyses confirm that multi-round game negotiations coupled with robust optimization maintain cooperation cohesion and risk tolerance under diverse hydro-economic scenarios. The findings underscore the utility of phased compensation, flexible reservoir releases, and multilateral funding to foster enduring cross-border cooperation in international river basins. © 2025 Elsevier Ltd

关键词： Risk perception

来源：评论

学校读者我要写书评

暂无评论

Assessing the Maturity of Blockchain-based Supply Chain Management and Logistics: Comprehensive Examination of the Current Situation

引用

Procedia computer Science 2025年 253卷 2458-2467页

作者： Umut Şener Murat Tahir Çaldağ Ebru Gökalp Graduate School of Informatics Middle East Technical University Ankara Türkiye Department of Technology and Knowledge Management Baskent University Ankara Türkiye Department of Computer Engineering Hacettepe University Ankara Türkiye

Blockchain technology, a transformative force in digital transactions, has emerged as a significant enabler of secure, transparent, and traceable operations within supply chain management and logistics (SCML). Its immutable nature facilitates enhanced management of SCML activities by seamlessly integrating information flow throughout the value chain, and substantially increasing data integrity and operational reliability. While organizations are attracted to these capabilities, they still lack comprehensive guidelines to fully benefit from blockchain in SCML. There is a fundamental need to assist organizations in assessing and enhancing their blockchain-based SCML capabilities. Maturity models (MMs) are tools that help organizations assess and improve their processes, providing step-by-step guidelines from basic to advanced levels of maturity. Therefore, this study aims to identify and analyze MMs focused on blockchain-based SCML. An extensive systematic literature review (SLR) was conducted, and thirteen relevant studies were identified and evaluated against a comprehensive set of quality criteria. The analysis revealed that none of the existing MMs fully satisfied all criteria, highlighting a significant gap in the field. This research provides a thorough systematic analysis of MMs for blockchain-based SCML, offering an in-depth view of their levels, dimensions, research focuses, domains, industry targets, and their quality evaluation. The significance of this study lies in identifying the necessity for an MM specifically designed for blockchain-based SCML, addressing a critical gap in the current literature. These findings contribute to the growing body of knowledge on blockchain integration in SCML and provide a foundation for future research.

关键词： Blockchain Supply Chain Management Logistics Logistics 4.0 Maturity Model Assessment

来源：评论

学校读者我要写书评

暂无评论

The Role of Digital Health Literacy and Socioeconomic Factors in Colorectal Cancer Screening: Machine Learning Analysis of HINTS Data

The Role of Digital Health Literacy and Socioeconomic Factor...

引用

18th International Joint Conference on Biomedical engineering Systems and Technologies, BIOSTEC 2025

作者： Kim, Sujin Dahal, Madhav Bhakta, Avinash Bae, Jihye Division of Biomedical Informatics University of Kentucky 725 Rose Street LexingtonKY United States Deptment of Electrical and Computer Engineering University of Kentucky 512 Administration Drive LexingtonKY United States Department of Surgery University of Kentucky 800 Rose Street LexingtonKY United States

While colorectal cancer (CRC) screening rates are on the rise, significant disparities persist, particularly among underserved populations, highlighting ongoing challenges in achieving equitable access to preventive care. This study utilizes machine learning models to analyze multi-year data from the Health Information National Trends Survey (HINTS), identifying critical factors influencing CRC screening adherence across three distinct time periods (2003–2008, 2011–2013, 2018–2020). Using Random Forest and Logistic Regression models, interpreted through Shapley Additive exPlanations values, we examine the impact of sociodemographic characteristics, digital health engagement, and digital literacy on CRC screening behaviors. Findings reveal that age, prior screening behavior, and digital literacy are key predictors;individuals with higher digital literacy, for example, exhibited a 22% higher likelihood of adhering to CRC screening guidelines. Age emerged as a dominant factor, with screening rates peaking at 43% in the 50–64 age group. These results suggest that interventions targeting digital health literacy and enhancing provider communication may effectively improve CRC screening rates among underserved populations. This study underscores the value of data-driven approaches in informing public health strategies to increase CRC screening adherence and reduce health disparities. © 2025 by SCITEPRESS – Science and Technology Publications, Lda.

关键词： Electronic health record

来源：评论

学校读者我要写书评

暂无评论

Partition-Based Cache Replacement to Manage Shared L2 Caches

引用

Chinese Journal of Electronics 2025年第3期23卷 464-467页

作者： Juan Fang Jing Wang Chengyan Li Zhicheng Yao Weimao Ke College of Computer Science Beijing University of Technology Beijing China College of Information Engineering Capital Normal University Beijing China College of Computing and Informatics Drexel University Chapel Hill NC USA

The Least recently used (LRU) replacement has been widely employed by on-chip shared caches. With the increase of cache associativity, the performance gap between LRU and theoretical optimal re-placement algorithm is significant, motivating us to design alternative replacement algorithms to improve cache performance. We propose a Partition-based cache replacement (PCR) algorithm to manage multi-core shared L2 caches. PCR takes into consideration recency and frequency information of cache access, and cache resource interference among multiple competing applications. PCR is effective to filter out the less frequently reused blocks through a dynamic insertion/promotion policy based on the target partitions, and ensures that some fraction of the working set is retained in the cache to contribute to more hits. Our evaluation shows that PCR improves performance of an 8-core system by an average of 11.3% over LRU-based cache.

关键词： Power demand Multicore processing Heuristic algorithms Education Interference Filtering algorithms Information filters Partitioning algorithms System-on-chip

来源：评论

学校读者我要写书评

暂无评论

Evaluating Beam Sweeping for AoA Estimation with an RIS Prototype: Indoor/Outdoor Field Trials

Evaluating Beam Sweeping for AoA Estimation with an RIS Prot...

引用

IEEE Conference on Wireless Communications and Networking

作者： Dimitris Vordonis Dimitris Kompostiotis Vassilis Paliouras George C. Alexandropoulos Florin Grec Electrical and Computer Engineering Department University of Patras Greece Department of Informatics and Telecommunications National and Kapodistrian University of Athens Greece European Space Agency (ESA) Noordwijk Netherlands

ISBN: (数字)9798350368369

ISBN: (纸本)9798350368376

Reconfigurable Intelligent Surfaces (RISs) have emerged as a promising technology to enhance wireless communication systems by enabling dynamic control over the propagation environment. However, practical experiments are crucial towards the validation of the theoretical potential of RISs while establishing their real-world applicability, especially since most studies rely on simplified models and lack comprehensive field trials. In this paper, we present an efficient method for configuring a 1-bit RIS prototype at sub-6 GHz, resulting in a codebook oriented for beam sweeping; an essential protocol for initial access and Angle of Arrival (AoA) estimation. The measured radiation patterns of the RIS validate the theoretical model, demonstrating consistency between the experimental results and the predicted beamforming behavior. Furthermore, we experimentally prove that RIS can alter channel properties and by harnessing the diversity it provides, we evaluate beam sweeping as an AoA estimation technique. Finally, we investigate the frequency selectivity of the RIS and propose an approach to address indoor challenges by leveraging the geometry of environment.

关键词： Geometry Wireless communication Protocols Prototypes Reconfigurable intelligent surfaces Size measurement Real-time systems Gain Wideband Antenna radiation patterns

来源：评论

学校读者我要写书评

暂无评论

Extremely Large Full Duplex MIMO for Simultaneous Downlink Communications and Monostatic Sensing at Sub-THz Frequencies

arXiv

引用

arXiv 2025年

作者： Alexandropoulos, George C. Gavras, Ioannis Department of Informatics and Telecommunications National and Kapodistrian University of Athens Athens16122 Greece Department of Electrical and Computer Engineering University of Illinois ChicagoIL60601 United States

The in-band Full Duplex (FD) technology is lately gaining attention as an enabler for the emerging paradigm of Integrated Sensing and Communications (ISAC), which envisions seamless integration of sensing mechanisms for unconnected entities into next generation wireless networks. In this paper, we present an FD Multiple-Input Multiple-Output (MIMO) system with extremely large antenna arrays at its transceiver module, which is optimized, considering two emerging analog beamforming architectures, for simultaneous DownLink (DL) communications and monostatic-type sensing intended at the sub-THz frequencies, with the latter operation relying on received reflections of the transmitted information-bearing signals. A novel optimization framework for the joint design of the analog and digital transmit beamforming, analog receive combining, and the digital canceler for the self-interference signal is devised with the objective to maximize the achievable DL rate, while meeting a predefined threshold for the position error bound for the unknown three-dimensional parameters of a passive target. Capitalizing on the distinctive features of the beamforming architectures with fully-connected networks of phase shifters and partially-connected arrays of metamaterials, two ISAC designs are presented. Our simulation results showcase the superiority of both proposed designs over state-of-the-art schemes, highlighting the role of various system parameters in the trade-off between the communication and sensing functionalities. © 2025, CC BY-NC-ND.

关键词： Metamaterial antennas

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：