检索结果-内蒙古大学图书馆

Building intelligence identification system via large language model watermarking: a survey and beyond

Artificial Intelligence Review 2025年第8期58卷 1-58页

作者： Wang, Xuhong Jiang, Haoyu Yu, Yi Yu, Jingru Lin, Yilun Yi, Ping Wang, Yingchun Qiao, Yu Li, Li Wang, Fei-Yue Shanghai Artificial Intelligence Laboratory Shanghai China School of Cyber Science and Engineering Shanghai Jiao Tong University Shanghai China Department of Automation BNRist Tsinghua University Beijing China The State Key Laboratory for Management and Control of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China

Large Language Models (LLMs) are increasingly integrated into diverse industries, posing substantial security risks due to unauthorized replication and misuse. To mitigate these concerns, robust identification mechanisms are widely acknowledged as an effective strategy. Identification systems for LLMs now rely heavily on watermarking technology to manage and protect intellectual property and ensure data security. However, previous studies have primarily concentrated on the basic principles of algorithms and lacked a comprehensive analysis of watermarking theory and practice from the perspective of intelligent identification. To bridge this gap, firstly, we explore how a robust identity recognition system can be effectively implemented and managed within LLMs by various participants using watermarking technology. Secondly, we propose a mathematical framework based on mutual information theory, which systematizes the identification process to achieve more precise and customized watermarking. Additionally, we present a comprehensive evaluation of performance metrics for LLM watermarking, reflecting participant preferences and advancing discussions on its identification applications. Lastly, we outline the existing challenges in current watermarking technologies and theoretical frameworks, and provide directional guidance to address these challenges. Our systematic classification and detailed exposition aim to enhance the comparison and evaluation of various methods, fostering further research and development toward a transparent, secure, and equitable LLM ecosystem.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Efficient Facial Expression Recognition Network Based on Spot-Adaptive Knowledge Distillation

Efficient Facial Expression Recognition Network Based on Spo...

引用

IEEE International Conference on Industrial Technology (ICIT)

作者： Xin Wen Zhen-Tao Liu Dan Chen School of Automation China University of Geosciences Wuhan China Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education

ISBN: (数字)9798331521950

ISBN: (纸本)9798331521967

Facial expression recognition plays a key role in promoting the development of comprehensive intelligence and building friendly human-computer interaction. Due to the interference of feature noise in expression data, the lightweight facial expression recognition model with fewer parameters is difficult to learn more expression features through simple training, which limits the improvement of its recognition performance. An efficient facial expression recognition network based on Spot-adaptive Knowledge Distillation is proposed in this paper. Inspired by VoVNetV2, the network designed in this paper is lightweightly improved using Depthwise Separable Convolution and the parameter-free SimAM attention mechanism, reducing the number of parameters to 0.21 M. To further improve the recognition accuracy of the model, Spot-adaptive Knowledge Distillation is employed to improve the characterization ability of the model. The recognition accuracies of the student network designed in this paper on the KDEF and RAF-DB datasets are 93.05% and 81.17% respectively after spot-adaptive distillation.

关键词： Knowledge engineering Training Human computer interaction Accuracy Face recognition Computational modeling Noise Neural networks Interference Data models

来源：评论

学校读者我要写书评

暂无评论

Multi-Level Features Fusion for Zero-Shot Object Pose Estimation

Multi-Level Features Fusion for Zero-Shot Object Pose Estima...

引用

IEEE International Conference on Industrial Technology (ICIT)

作者： Nanxin Huang Chi Xu School of Automation China University of Geosciences Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education Wuhan China

ISBN: (数字)9798331521950

ISBN: (纸本)9798331521967

Driven by advancements in industrial production and artificial intelligence, the need for pose estimation of new ob-jects in areas like robotic manipulation and virtual reality is increasing. We introduce a zero-shot object pose estimation approach that identifies the poses of objects excluded from the training dataset, removing the requirement for re-modeling. The method is built around a multi-level features fusion framework de-signed to enhance generalization. First, a trainable feature extraction module filters and selects multi-level features extracted by the backbone network. Unlike traditional convolutional ker-nels, we incorporate a dynamic convolution kernel to enhance the feature extraction capability. Second, in the feature fusion module, we adopt a dynamic weight generation strategy to perform weighted fusion of multi-level features. This method enhances template matching by effectively describing similarities between unseen objects (those absent from the training set) and templates, leveraging robust and adaptive feature representations to narrow the gap with seen objects. Experimental results demonstrate that our approach achieves state-of-the-art performance on two popu-lar benchmark datasets, LineMod and LineMod-Occlusion, proves that our method has better generalization than previous models.

关键词： Training Solid modeling Convolution Service robots Pose estimation Zero shot learning Virtual reality Benchmark testing Feature extraction Robustness

来源：评论

学校读者我要写书评

暂无评论

Intrinsic-feature-guided 3D Object Detection

arXiv

引用

arXiv 2025年

作者： Zhang, Wanjing Wang, Chenxing School of Automation Southeast University 2# Sipailou Xuanwu Nanjing210096 China Key laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Southeast University Nanjing210096 China

LiDAR-based 3D object detection is essential for autonomous driving systems. However, LiDAR point clouds may appear to have sparsity, uneven distribution, and incomplete structures, significantly limiting the detection performance. In road driving environments, target objects referring to vehicles, pedestrians and cyclists are well-suited for enhancing representation through the complete template guidance, considering their grid and topological structures. Therefore, this paper presents an intrinsic-feature-guided 3D object detection method based on a template-assisted feature enhancement module, which extracts intrinsic features from relatively generalized templates and provides rich structural information for foreground objects. Furthermore, a proposal-level contrastive learning mechanism is designed to enhance the feature differences between foreground and background objects. The proposed modules can act as plug-and-play components and improve the performance of multiple existing methods. Extensive experiments illustrate that the proposed method achieves the highly competitive detection results. Code will be available at https://***/zhangwanjingjj/***. Copyright © 2025, The Authors. All rights reserved.

关键词： Object recognition

来源：评论

学校读者我要写书评

暂无评论

Advances in Robust Federated Learning: A Survey with Heterogeneity Considerations

引用

IEEE Transactions on Big Data 2025年第03期11卷 1548-1567页

作者： Chen, Chuan Liao, Tianchi Deng, Xiaojun Wu, Zihou Huang, Sheng Zheng, Zibin Sun Yat-sen University School of Computer Science and Engineering Guangzhou510006 China Guangdong Key Laboratory of Big Data Analysis and Processing China Sun Yat-sen University School of Software Engineering Zhuhai519000 China Sun Yat-sen University School of Systems Science and Engineering Guangzhou510006 China Zhuhai Key Laboratory of Trusted Large Language Models China

In the field of heterogeneous federated learning (FL), the key challenge is to efficiently and collaboratively train models across multiple clients with different data distributions, model structures, task objectives, computational capabilities, and communication resources. This diversity leads to significant heterogeneity, which increases the complexity of model training. In this paper, we first outline the basic concepts of heterogeneous FL and summarize the research challenges in FL in terms of five aspects: data, model, task, device and communication. In addition, we explore how existing state-of-the-art approaches cope with the heterogeneity of FL, and categorize and review these approaches at three different levels: data-level, model-level, and architecture-level. Subsequently, the paper extensively discusses privacy-preserving strategies in heterogeneous FL environments. Finally, the paper discusses current open issues and directions for future research, aiming to promote the further development of heterogeneous FL. © 2015 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

EK-Net++: Real-time scene text detection with expand kernel distance and Epoch Adaptive Weight

引用

Expert systems with Applications 2025年 267卷

作者： Zhu, Boyuan Chen, Xi Tang, Quan Chen, C.L. Philip Liu, Fagui Computer Science and Engineering South China University of Technology Guangzhou510006 China Cyberspace Security Research Center Peng Cheng Laboratory Shenzhen518066 China State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing100080 China

Recently, scene text detection has received significant attention due to its wide applications. Accurate detection in complex scenes of multiple scales, orientations, and curvature remains a challenge. Component-based methods have emerged as the dominant approach for offering a more flexible way to achieve superior performance. However, Component-based methods employ the Vatti clipping algorithm for multi-instance training to handle the difficulty of arbitrary-shaped text. Yet we identify several bias results from these approaches called the "shrinked kernel" leading to decreased accuracy. Additionally, almost all text detection methods fail to consider the characteristics of different datasets and directly assign fixed loss weights to multi-instances. To address these limitations, we propose a new approach named Expand Kernel Network (EK-Net++), which presents the expand kernel distance map to compensate for the previous deficiency and incorporates three-stage regression to complete integral instance detection. Furthermore, we design the Epoch Adaptive Weight algorithm which dynamically distinguishes the characteristics of different datasets. EK-Net++ not only achieves precise positioning of arbitrary-shaped text but also strikes a balance between performance and speed. Evaluation results show that EK-Net++ attains state-of-the-art or competitive performance compared to other advanced methods, e.g., F-measure of 86.72% at 35.42 FPS on ICDAR 2015, F-measure of 85.83% at 40.13 FPS on CTW1500. © 2024 Elsevier Ltd

关键词： Epoch Adaptive Weight Expand kernel distance Real-time Scene text detection

来源：评论

学校读者我要写书评

暂无评论

Breaking the SSL-AL Barrier: A Synergistic Semi-Supervised Active Learning Framework for 3D Object Detection

arXiv

引用

arXiv 2025年

作者： Wang, Zengran Zhang, Yanan Chen, Jiaxin Huang, Di State Key Laboratory of Complex and Critical Software Environment Beihang University Beijing100191 China School of Computer Science and Engineering Beihang University Beijing100191 China School of Computer Science and Information Engineering Hefei University of Technology Hefei230601 China

To address the annotation burden in LiDAR-based 3D object detection, active learning (AL) methods offer a promising solution. However, traditional active learning approaches solely rely on a small amount of labeled data to train an initial model for data selection, overlooking the potential of leveraging the abundance of unlabeled data. Recently, attempts to integrate semi-supervised learning (SSL) into AL with the goal of leveraging unlabeled data have faced challenges in effectively resolving the conflict between the two paradigms, resulting in less satisfactory performance. To tackle this conflict, we propose a Synergistic Semi-Supervised Active Learning framework, dubbed as S-SSAL. Specifically, from the perspective of SSL, we propose a Collaborative PseudoScene Pre-training (CPSP) method that effectively learns from unlabeled data without introducing adverse effects. From the perspective of AL, we design a Collaborative Active Learning (CAL) method, which complements the uncertainty and diversity methods by model cascading. This allows us to fully exploit the potential of the CPSP pre-trained model. Extensive experiments conducted on KITTI and Waymo demonstrate the effectiveness of our S-SSAL framework. Notably, on the KITTI dataset, utilizing only 2% labeled data, S-SSAL can achieve performance comparable to models trained on the full dataset. The code has been released at https://***/LandDreamer/S SSAL. Copyright © 2025, The Authors. All rights reserved.

关键词： Active learning

来源：评论

学校读者我要写书评

暂无评论

Negative-pressure soft pneumatic actuators enabled by asymmetrically distributed bending units

引用

Extreme Mechanics Letters 2025年 77卷

作者： Fu, Zeyu Chen, Tianyu Zou, Haoran Zhu, Zhiwei Deng, Zichen Wang, Yifan School of Mechanics Civil Engineering and Architecture Northwestern Polytechnical University Xi'an710072 China MIIT Key Laboratory of Dynamics and Control of Complex Systems Northwestern Polytechnical University Xi'an710072 China School of Mechanical and Aerospace Engineering Nanyang Technological University Singapore 639798 Singapore State Key Laboratory of Mechanical Behavior and System Safety of Traffic Engineering Structures Shijiazhuang Tiedao University Shijiazhuang050043 China

Soft pneumatic actuators are highly flexible and can adapt to complex environments, attracting significant attention for their ability to operate in settings where rigid actuators cannot. Most existing soft pneumatic actuators are powered by positive pressure, which has the disadvantage of volume increase during operation and risk of exploding under large pressures. In this study, we designed a new type of soft actuator by inducing asymmetric bending deformation throughout a holey structure under negative pressure. Using finite element analysis and a theoretical model, we studied the effects of various geometric parameters on the structure's bending behavior. We found that the displacement of the circle's center from the central axis is a parameter with the most significant impact on the bending behavior of the structure. By designing circular holes with offset from the central axis, the structure can bend toward the offset direction during subsequent bending deformation. Based on this deformation mechanism, we designed a soft gripper that bends upon actuation and effectively picks up objects. We further designed a bi-directional gripper that bends towards both sides and a circular gripper that contracts towards the center upon evacuation to hold objects. These soft grippers’ grasping capabilities are validated by experimental tests. Finally, by arranging the holes in a wavy pattern within the structure, we created a crawling robot that moves forward through cyclic negative pressure actuation. These findings highlight the potential of leveraging bending behaviors in asymmetrically distributed holey structures for the development of negative-pressure driven soft actuators and robots. © 2025 Elsevier Ltd

关键词： Pneumatic actuators

来源：评论

学校读者我要写书评

暂无评论

Finite-time Bipartite Synchronization of Homogeneous and Heterogeneous Multiple Agents with Input Saturation: A TVRE-based Gain Approach

引用

IEEE Transactions on Automatic Control 2025年

作者： Zhou, Yuan Liu, Yongfang Zhao, Yu Li, Zhongkui Hong Kong Polytechnic University Department of Applied Mathematics Hung Hom Hong Kong Northwestern Polytechnical University School of Automation Xi'an China Peking University State Key Laboratory for Turbulence and Complex Systems Department of Mechanics and Engineering Science College of Engineering Beijing100871 China

Pursuing faster convergence rates and smaller input magnitudes seem to be two conflicting goals in studying multi-agent systems. To give a trade-off between the two, this paper focuses on the bipartite synchronization problems over signed topologies and aims to achieve finite-time control for general linear agents subject to input saturation constraints. Firstly, this paper considers homogeneous agents and presents a class of bipartite synchronization protocols with saturation constraint, which exploits the solution of the time-varying Riccati equation (TVRE) to design the control gain. Then, a time-varying parameter scheduler is tactically designed for TVRE and achieves finite-time bipartite synchronization. Note that the design uses the solution computed online and brings a bit of conservatism in determining the settling time. So, for heterogeneous agents, this paper constructs a modified parameter scheduler computed off-line to reduce the conservatism. A class of finite-time bipartite synchronization generators and generator-based finite-time protocols are proposed. It shows that, in both designs, the control input subjects to the bound saturation during convergence even if the gain escapes to infinity towards the settling time. Moreover, the trade-off among the finite convergence time, the saturation bound of the input, and the initial domain are analyzed explicitly in theory. Finally, two simulations verify the validity of the theoretical results. © 1963-2012 IEEE.

关键词： Synchronization

来源：评论

学校读者我要写书评

暂无评论

Design, Dynamic Modeling and Control of a 2-DOF Robotic Wrist Actuated by Twisted and Coiled Actuators

arXiv

引用

arXiv 2025年

作者： Zhang, Yunsong Zhou, Xinyu Zhang, Feitian Department of Advanced Manufacturing and Robotics College of Engineering State Key Laboratory of Turbulence and Complex Systems Peking University Beijing100871 China Department of Electrical and Computer Engineering Michigan State University East LansingMI48824 United States

Robotic wrists play a pivotal role in the functionality of industrial manipulators and humanoid robots, facilitating manipulation and grasping tasks. In recent years, there has been a growing interest in integrating artificial muscle-driven actuators for robotic wrists, driven by advancements in technology offering high energy density, lightweight construction, and compact designs. However, in the study of robotic wrists driven by artificial muscles, dynamic model-based controllers are often overlooked, despite their critical importance for motion analysis and dynamic control of robots. This paper presents a novel design of a two-degree-of-freedom (2-DOF) robotic wrist driven by twisted and coiled actuators (TCA) utilizing a parallel mechanism with a 3RRRR configuration. The proposed robotic wrist is expected to feature lightweight structures and superior motion performance while mitigating friction issues. The Lagrangian dynamic model of the wrist is established, along with a nonlinear model predictive controller (NMPC) designed for trajectory tracking tasks. A prototype of the robotic wrist is developed, and extensive experiments are conducted to validate its superior motion performance and the proposed dynamic model. Subsequently, extensive comparative experiments between NMPC and PID controller were conducted under various operating conditions. The experimental results demonstrate the effectiveness and robustness of the dynamic model-based controller in the motion control of TCA-driven robotic wrists. Copyright © 2025, The Authors. All rights reserved.

关键词： Industrial manipulators

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：