检索结果-内蒙古大学图书馆

Semantic-Aware Pseudo-Labeling for Unsupervised Meta-Learning

IEEE Transactions on Pattern Analysis and Machine Intelligence 2025年 PP卷 PP页

作者： Ouyang, Tianran Dong, Xingping Ye, Mang Du, Bo Shao, Ling Shen, Jianbing Wuhan University School of Computer Science National Engineering Research Center for Multimedia Software Institute of Artificial Intelligence Hubei Key Laboratory of Multimedia and Network Communication Engineering Wuhan430072 China University of Chinese Academy of Sciences UCAS-Terminus AI Lab Beijing China University of Macau State Key Laboratory of Internet of Things for Smart City Department of Computer and Information Science China

In unsupervised meta-learning, the clustering-based pseudo-labeling approach is an attractive framework, since it is model-agnostic, allowing it to synergize with supervised algorithms to learn from unlabeled data. However, the pseudo-labels suffer from clustering noise and semantic chaos problems, further impacting the effectiveness of meta-learning. In this paper, we analyze and optimize the pseudo-labeling process, including encoding and clustering, aiming to generate semantic- like pseudo-labels to narrow the gap between unsupervised and supervised meta-learning. Firstly, during the encoding, we observe that the embedding space of existing methods lacks clustering-friendly properties, which is the primary reason for clustering noise. To address this issue, we minimize the inter-to-intra-class similarity ratio to generate clustering-friendly embedding features and validate our approach through comprehensive experiments. Then, during the clustering, we find that the semantic quality of pseudo-labels is not adequately controlled, resulting in semantic chaos of pseudo-labels. We propose a semantic-stability index to measure the semantic quality of pseudo-labels quantitatively. Based on this index, we propose the Semantic-aware Pseudo-label Reassignment mechanism to generate semantic- like pseudo-labels for all samples. Our approach is model-agnostic and can easily be integrated into existing supervised methods. To demonstrate its generalization ability, we integrate it into two representative algorithms: MAML and EP. The results on three main few-shot benchmarks clearly show that the proposed method achieves significant improvement compared to state-of-the-art models. Notably, our approach also outperforms the corresponding supervised method in three tasks. The source code will be made publicly available. ©1979-2012 IEEE.

关键词： Latent semantic analysis

来源：评论

学校读者我要写书评

暂无评论

Sensing-communication-Computation Integration for Federated Edge Learning with Controllable Model Dropout

引用

IEEE Internet of Things Journal 2025年

作者： Jiao, Xiang Zhu, Guangxu Jiang, Wei Chen, Li Luo, Wu Wen, Dingzhu State Key Laboratory of Advanced Optical Communication Systems and Networks School of Electronics Peking University Beijing China Shenzhen Research Institute of Big Data Shenzhen China The State Key Laboratory of Advanced Optical Communication Systems and Networks School of Electronics Peking University Beijing China The CAS Key Laboratory of Wireless Optical Communication University of Science and Technology of China Hefei China Network Intelligence Center School of Information Science and Technology ShanghaiTech University Shanghai China

FEderated Edge Learning (FEEL) is an advanced paradigm in edge artificial intelligence, enabling privacy-preserving collaborative model training through periodic communication between edge devices and a central server. FEEL involves three key processes: sensing, computation, and communication for data acquisition, processing, and exchange, respectively. Due to limited system resources, optimizing each process individually may lead to suboptimal learning performance. This challenge has sparked research into Integrated Sensing-Computation-communication (ISCC) design for enhanced FEEL. While previous work has optimized general learning parameters such as batch size and computing frequency, there is a lack of customized designs considering the neural network architecture as an optimizable variable in ISCC for FEEL. To close this gap, we introduce a novel design where each device generates a sub-model through controllable weight dropout, adding flexibility by directly manipulating the learning process and reducing computation and communication overhead. To guide ISCC resource allocation in this new setting, we present a comprehensive convergence analysis, revealing the tight coupling of sensing, computation, and communication across devices and their impact on FEEL convergence. Building on these theoretical insights, we formulate an ISCC problem aiming to maximize the FEEL convergence rate through joint optimization of variables such as batch size, sensing power, dropout rate, and communication power. This non-convex problem is decomposed into two sub-problems via alternating optimization: one controls batch size using a sorting algorithm, while the other focuses on ISCC device parameters, transformable into a convex problem solved by successive convex approximation. Extensive experiments using human motion recognition datasets demonstrate the superiority of the proposed design over baseline schemes. © 2014 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Contextual Graph Reconstruction and Emotional Variation Learning for Conversational Emotion Recognition

引用

IEEE Transactions on Affective Computing 2025年

作者： Rao, Yujing Cao, Min Ye, Mang Wuhan University National Engineering Research Center for Multimedia Software Institute of Artificial Intelligence Hubei Key Laboratory of Multimedia and Network Communication Engineering School of Computer Science Wuhan 430072 China Soochow University School of Computer Science and Technology Suzhou 215006 China

Conversational Emotion Recognition (CER) significantly benefits from the integration of multiple modalities. However, real-world scenarios are often plagued by hardware malfunctions and network failures that lead to missing modalities and incomplete emotional representations. Existing methods primarily focus on modeling inter-modal relationships within isolated utterances to generate missing data, so they do not adequately capture conversational context and dynamic emotional evolution. In conversations, emotional expressions are inherently context-dependent and dynamic. Insufficient modeling of these properties can substantially degrade performance. To address these challenges, we propose Contextual Graph Reconstruction and Emotional Variation Learning (CGR-EVL). Our approach flexibly constructs an utterance graph with diverse coverage based on speaker activity and modality availability, thereby capturing more contextual information. To enhance the handling of missing modalities, we further integrate temporal relations through a relational graph convolutional network to reconstruct missing features and introduce a speaker emotion-aware constraint to ensure emotional coherence. Additionally, we propose the concept of emotional entropy to quantify variation patterns and develop a novel loss function that aligns predicted and actual variations. Experiments on the IEMOCAP and MELD datasets show that CGR-EVL outperforms state-of-the-art methods, particularly under conditions of incomplete modalities. © 2010-2012 IEEE.

关键词： Emotional variation graph neural network modality reconstruction multimodal emotion recognition

来源：评论

学校读者我要写书评

暂无评论

communication-Efficient Personalized Distributed Learning with Data and Node Heterogeneity

引用

IEEE Transactions on Cognitive communications and networking 2025年

作者： Tian, Zhuojun Zhang, Zhaoyang Li, Yiwei Bennis, Mehdi Zhejiang University College of Information Science and Electronic Engineering Hangzhou310027 China University of Oulu Center for Wireless Communications Oulu90014 Finland State Key Laboratory of Industrial Control Technology Hangzhou310027 China Zhejiang Provincial Key Laboratory of Multimodal Communication Networks and Intelligent Information Processing Hangzhou310027 China Xiamen University of Technology Fujian Key Laboratory of Communication Network and Information Processing Xiamen361024 China

To jointly tackle the challenges of data and node heterogeneity in decentralized learning, we propose a distributed strong lottery ticket hypothesis (DSLTH), based on which a communication-efficient personalized learning algorithm is developed. In the proposed method, each local model is represented as the ‌Hadamard product of global real-valued parameters and a personalized binary mask for pruning. The local model is learned by updating and fusing the personalized binary masks while the real-valued parameters are fixed among different agents. To further reduce the complexity of hardware implementation, we incorporate a group sparse regularization term in the loss function, enabling the learned local model to achieve structured sparsity. Then, a binary mask aggregation algorithm is designed by introducing an intermediate aggregation tensor and adding a personalized fine-tuning step in each iteration, which constrains model updates towards the local data distribution. The proposed method effectively leverages the relativity among agents while meeting personalized requirements in heterogeneous node conditions. We also provide a theoretical proof for the DSLTH, establishing it as the foundation of the proposed method. Numerical simulations confirm the validity of the DSLTH and demonstrate the effectiveness of the proposed algorithm. © 2015 IEEE.

关键词： Iterative methods

来源：评论

学校读者我要写书评

暂无评论

Multi-Modal Sensing and Lighting Estimation for PoE-Enabled Intelligent and Energy-Efficient Lighting networks

引用

IEEE Internet of Things Journal 2025年

作者： Tang, Liwei Xu, Yuanyuan Liu, Xiaodong Wang, Yuhao Wang, Jinyuan Shao, Huilong Nanchang University School of Information Engineering Nanchang330031 China Nanchang Laboratory Nanchang330224 China Nanjing University of Posts and Telecommunications Key Laboratory of Broadband Wireless Communication and Sensor Network Technology Nanjing210003 China

The application of power over Ethernet (PoE)-enabled lighting has emerged as a groundbreaking technology with the potential to achieve cooperative control and create an energy-saving lighting network. However, fine-grained user-side lighting environment perception is crucial to improve energy efficiency and user experience. To tackle this issue, a PoE-enabled intelligent lighting network with multi-modal sensing is proposed. Specifically, a multi-modal sensor is designed and installed on the side of the luminaire to provide both fine-grained lighting environment monitor and dimming. Then, a multilayer perceptron (MLP)-based estimation scheme is proposed to estimate the illumination and color temperature information around the user activity area. Moreover, an experimental lighting network is constructed and deployed in a classroom to measure practical lighting dataset. The proposed scheme effectively estimates the lighting quality on the user side based on the measurement data. Furthermore, based on the estimated information and the collaborative control capabilities of PoE, the operating current of the luminaire is updated in real time, and thus the ambient light resource is fully utilized. Finally, the proposed lighting network achieves a 39.99% energy saving. © 2014 IEEE.

关键词： Lighting fixtures

来源：评论

学校读者我要写书评

暂无评论

Design of campus security management and control based on ZigBee 4

Design of campus security management and control based on Zi...

引用

4th International Conference on network communication and Information Security, ICNCIS 2024

作者： Liu, Yulong Wang, Mingjun Ma, Xiaopan School of Automation an Information Engineering Xi’an University of Technology Shaanxi Xi’an710048 China Xi'an Key Laboratory of Wireless Optical Communication and Network Research Shaanxi Xi’an710048 China School of Physics and Telecommunications Engineering Shanxi University of Technology Shaanxi Hanzhong723001 China China Mobile Tietong Co. Ltd. Yulin Branch Shaanxi Yulin719000 China

ISBN: (数字)9781510688261

ISBN: (纸本)9781510688254

In view of the great significance of the Internet of Things technology to the construction of security management and control in chemical parks, this paper designs a network transmission technology of security management and control equipment based on ZigBee, which realizes the data transmission and control of various monitoring terminals such as cameras and sensors. At the same time, the security control of the park is designed in detail in three parts: basic security management, monitoring and early warning of major hazard sources and video monitoring. The results show that the security management and control system of the park effectively solves the existing problems of security risks in the park from three aspects: basic security management, monitoring and early warning of major hazard sources and video monitoring, improves the efficiency and level of supervision of the park management committee, and provides factual basis and construction experience for risk management, early warning and decision-making of other chemical parks in China. © 2025 SPIE.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

ROMA: ROtary and Movable Antenna

arXiv

引用

arXiv 2025年

作者： Zhang, Jiayi Yi, Wenhui Xu, Bokai Wang, Zhe Xiao, Huahua Ai, Bo State Key Laboratory of Advanced Rail Autonomous Operation China School of Electronics and Information Engineering Beijing Jiaotong University Beijing100044 China ZTE Corporation State Key Laboratory of Mobile Network and Mobile Multimedia Technology Shenzhen518055 China

The rotary and movable antenna (ROMA) architecture represents a next-generation multi-antenna technology that enables flexible adjustment of antenna position and array rotation angles of the transceiver. In this letter, we propose a ROMA-aided multi-user MIMO communication system to enhance the efficiency and reliability of system transmissions fully. By deploying ROMA panels at both the transmitter and receiver sides, and jointly optimizing the three-dimensional (3D) rotation angles of each ROMA panel and the relative positions of antenna elements based on the spatial distribution of users and channel state information (CSI), we can achieve the objective of maximizing the average spectral efficiency (SE). Subsequently, we conduct a detailed analysis of the average SE performance of the system under the consideration of maximum ratio (MR) precoding. Due to the non-convexity of the optimization problem in the ROMA multi-user MIMO system, we propose an efficient solution based on an alternating optimization (AO) algorithm. Finally, simulation results demonstrate that the AO-based ROMA architecture can significantly improve the average SE. Furthermore, the performance improvement becomes more pronounced as the movable region and the transmission power increase. Copyright © 2025, The Authors. All rights reserved.

关键词： Convex optimization

来源：评论

学校读者我要写书评

暂无评论

HIAWSC: An Immune Algorithm Based Heuristic Web Service Composition Framework

引用

Chinese Journal of Electronics 2025年第3期23卷 579-585页

作者： Jiuyun Xu Reiff-Marganiec Stephan School of Computer and Communication Engineering China University of Petroleum Qingdao China Network and Switch Technology Key Laboratory Beijing University of Posts and Telecommunication Beijing China Department of Computer Science University of Leicester Leicester UK

The introduction of of web services has led to web service composition being a focus of many researchers. Composing web services using workflows is seen as the most realistic method from an industrial view-point. Amongst other method, the use of natural computing methods has been proposed previously to automate web service composition. The need for a fast response when computing the most suitable sequence of services is addressed in this paper. In particular, we propose a novel heuristic immune algorithm with an efficient encoding and mutation method. The algorithm involves two steps: an immune selection operation, which is maintaining antibody population diversity and the clonal selection. The use of a vaccine during the evolution provides heuristic information that accelerates the convergence. Our experimental results illustrate that the proposed heuristic immune algorithm is very effective in improving the convergence speed. We also provide a schema analysis for this method.

关键词： Web services Heuristic algorithms Antibodies Encoding Vaccines Convergence

来源：评论

学校读者我要写书评

暂无评论

Evaluating Radio Interfaces Towards 6G: What’s New and What’s Different

引用

IEEE network 2025年

作者： Liu, Ruiqi Zhang, Leyi Zou, Minqiang Parolari, Sergio Tian, Li State Key Laboratory of Mobile Network and Mobile Multimedia Technology Shenzhen China ZTE Corporation Wireless and Computing Research Institute Shenzhen China ZTE Corporation Technology Planning Department Shenzhen China ZTE Italia Strategy and Architecture Team Milan Italy

The International Telecommunication Union (ITU) has set forth the timetable to develop the 6th generation (6G) terrestrial networks, namely, the International Mobile Telecommunications towards 2030 and beyond (IMT-2030). Due to the fact that 6G is envisaged to support a much wider variety of usage scenarios and use cases, evaluations of candidate 6G radio interface technologies (RITs) are likely to be more challenging than those for the previous generation. Novel technical performance requirements (TPRs) as well as proper evaluation methodologies are essential to be defined to characterize the advanced capabilities of 6G. In this article, comprehensive investigations of potential new 6G TPRs as well as evaluation methodologies are provided. © 1986-2012 IEEE.

关键词： Radio

来源：评论

学校读者我要写书评

暂无评论

Cross-Modal Haptic Compression Inspired by Embodied AI for Haptic communications

引用

IEEE Transactions on multimedia 2025年

作者： Lu, Hang Tan, Xinmeng Chen, Mingkai Zhang, Zhe Zhang, Xuguang Chen, Jianxin Wei, Xin Zhao, Tiesong Nanjing University of Posts and Telecommunications Key Laboratory of Broadband Wireless Communication and Sensor Network Technology China Fuzhou University Fujian Key Laboratory for Intelligent Processing and Wireless Transmission of Media Information College of Physics and Information Engineering Fuzhou350108 China Fuzhou University Fujian Science & Technology Innovation Laboratory for Optoelectronic Information Fuzhou350108 China

Haptic data compression has gradually become a key issue for emerging real-time haptic communications in Tactile Internet (TI). However, it is challenging to achieve a tradeoff between high perceptual quality and compression ratio in haptic data compression scheme. Inspired by the perspective of embodied AI, we propose a cross-modal haptic compression scheme for haptic communications to improve the perception quality on TI devices in this paper. Since multimodal fusion is routinely employed to improve the ability of system in cognition, we assume that haptic codec is guided by visual semantics to optimize parameter settings in the coding process. We first design a multi-dimensional tactile feature fusion network (MTFFN) relying on multi-head attention mechanism. The MTFFN extracts the multi-dimensional features from the material surface and maps them to infer the coding parameters. Secondly, we provide second-order difference and linear interpolation to establish an criterion for the determination of optimal codec parameters, which are customized by the material categories so as to give high robustness. Finally, the simulation results reveal that our compression scheme can efficiently make a personalized codec procedure for different materials, obtaining more than 17% improvement in terms of compression ratio with high perceptual quality at the same time. © 2025 IEEE.

关键词： Data compression ratio

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：