检索结果-内蒙古大学图书馆

Proceedings of the 41st International Conference on Machine Learning

作者： Le Yu Xinde Li Pengfei Zhang Zhentong Zhang Fir Dunkin School of Automation Southeast University Nanjing China and Nanjing Center for Applied Mathematics Nanjing China and Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Southeast University Nanjing China School of Automation Southeast University Nanjing China and Nanjing Center for Applied Mathematics Nanjing China and Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Southeast University Nanjing China and Southeast University Shenzhen Research Institute Shenzhen China School of Automation Southeast University Nanjing China and Nanjing Center for Applied Mathematics Nanjing China

Model-Agnostic Meta-Learning (MAML) and its variants have shown remarkable performance in scenarios characterized by a scarcity of labeled data during the training phase of machine learning models. Despite these successes, MAMLbased approaches encounter significant challenges when there is a substantial discrepancy in the distribution of training and testing tasks, resulting in inefficient learning and limited generalization across domains. Inspired by classical proportional-integral-derivative (PID) control theory, this study introduces a Layer-Adaptive PID (LA-PID) Optimizer, a MAML-based optimizer that employs efficient parameter optimization methods to dynamically adjust task-specific PID control gains at each layer of the network, conducting a first-principles analysis of optimal convergence conditions. A series of experiments conducted on four standard benchmark datasets demonstrate the efficacy of the LA-PID optimizer, indicating that LA-PID achieves state-of-the-art performance in few-shot classification and cross-domain tasks, accomplishing these objectives with fewer training steps. Code is available on https://***/yuguopin/LA-PID.

关键词：

来源：评论

学校读者我要写书评

暂无评论

AI4S Based on DeSci: Reference Model and Research Issues

AI4S Based on DeSci: Reference Model and Research Issues

引用

IEEE International Conference on systems, Man and Cybernetics

作者： Wenwen Ding Juanjuan Li Rui Qin Sangtian Guan Fei-Yue Wang Faculty of Innovation Engineering Macau University of Science and Technology Macau China The State Key Laboratory for Management and Control of Complex Systems The State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of Automation Chinese Academy of Sciences Beijing China The State Key Laboratory for Management and Control of Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China

The rise of Artificial Intelligence for Science (AI4S) has highlighted the importance and urgency of ensuring open-ness, fairness, impartiality, diversity, and sustainability in scientific systems. Existing scientific systems, referred to as Centralized Science (CeSci), are built on centralized organizational structures and top-down institutional frameworks, which are lagging behind the development and practical requirements of AI4S. To address these limitations, AI4S needs to embrace a new scientific organizational and operational paradigm, namely Decentralized Science (DeSci). It can provide strong support to AI4S via effectively addressing issues such as information silos, biases, unfair distribution, and monopolies and promoting multidisciplinary, interdisciplinary, and trans disciplinary cooperation in science. Based on these considerations, this paper presents the framework of AI4S based on DeSci and explores its potential application scenarios and research issues. The research can provide effective guidance for the development of scientific systems.

关键词：

来源：评论

学校读者我要写书评

暂无评论

3D Reconstruction of Weakly Textured Objects Based On Denoising Polarized Images 41

3D Reconstruction of Weakly Textured Objects Based On Denois...

引用

第41届中国控制会议

作者： Youai Wang Yi Liu 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 Eduction CRRC Zhuzhou Locomotive Co. Ltd.

ISBN: (数字)9789887581536

ISBN: (纸本)9781665482561

Traditional 3D visual reconstruction technology based on feature extraction is difficult to estimate the surface depth information of a smooth object with a single texture,while the polarization 3 D reconstruction technology can recover the surface normal information from the image pixel by pixel,which has the characteristics of high precision,low texture dependence and strong material ***,the normals recovered from the polarization information have the problem of multi-value ambiguity,which needs to be corrected before the accurate normal information can be *** the same time,the image collected by the polarization camera contains a lot of noise,which will greatly affect the accuracy of reconstruction *** this paper,the characteristics of polarization information are used to denoise the polarization image,and the reference normal information is used as the prior information for correcting the polarization ***,the degree of polarization and angle of polarization are calculated based on the polarized image,then the two are fused and denoised by the Block Matching threedimensional(BM3D) method,and then the multi-valued ambiguity of polarization normal is corrected according to the reference normal ***,the corrected polarization normal is integrated to obtain a high-precision 3D shape of the object.

关键词： Polarization Denoising 3D Reconstruction Weakly Textured Objects Normal Multi-value Correction

来源：评论

学校读者我要写书评

暂无评论

Asymptotic Stabilization of Time-Varying Delay System with Actuator Saturation via Multi-dimensional Taylor Network

Asymptotic Stabilization of Time-Varying Delay System with A...

引用

International Conference on Robotics and Automation engineering (ICRAE)

作者： Guo-Biao Wang Hong-Sen Yan School of Automation Southeast University Nanjing Jiangsu China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing Jiangsu China

This paper addresses the stabilization problem of time-varying delay systems subject to actuator saturation. A novel class of Lyapunov functional with a nonlinear metric is developed to guarantee that there exists a Lyapunov-Razumikhin function that strictly decreases for the controlled system. Such a nonlinear metric is parameterized by a multi-dimensional Taylor network with concise topology, which brings into the high-order dynamic information and guarantees real-time performance. A sum-of-squares programming approach based on the designed Lyapunov functional is formulated to deduce the stability criteria, which maximizes the estimate of the region of attraction and ensures the uniform asymptotic stability of the closed-loop system theoretically. A numerical example demonstrates the effectiveness of the proposed stabilization control scheme.

关键词： measurement Actuators Asymptotic stability Delay systems Stability criteria Programming Real-time systems

来源：评论

学校读者我要写书评

暂无评论

A Robot Coordinate measurement System Based on Pull Wire Sensor and Its Parameter Identification Method 12

A Robot Coordinate Measurement System Based on Pull Wire Sen...

引用

12th International Conference on CYBER Technology in Automation, control, and Intelligent systems, CYBER 2022

作者： Gao, Shuang Ma, Kaiwei Gao, Yang Shen, Xin Yang, Mingxing Xu, Fengyu College of Automation and College of Artificial Intelligence Nanjing University of Posts and Telecommunications Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Southeast University Jiangsu Nanjing China Anhui Province Key Laboratory of Special Heavy Load Robot Anhui Maanshan China College of Automation and College of Artificial Intelligence Nanjing University of Posts and Telecommunications Jiangsu Nanjing China

ISBN: (数字)9781665472678

ISBN: (纸本)9781665472678

Aiming at the problems of large robot error and difficult coordinate detection, a robot coordinate measurement system and parameter identification method based on pull wire sensor are proposed. Firstly, a robot coordinate measurement system based on pull wire sensor is established. Secondly, Levenberg-Marquarelt algorithm is used to solve the kinematics parameters of the robot. Finally, the neural network of the whole connection layer is used to predict the optimal compensation amount for secondary compensation. In order to verify the above theory, the robot coordinate measurement system is built and verified on the five degree of freedom robot experimental platform. The results show that the robot coordinate measurement system can measure the robot coordinates quickly and accurately. The secondary compensation method controls the accuracy of the robot to about 2mm, and the optimal error is controlled within 0.012mm. The absolute positioning accuracy of the robot is effectively improved. © 2022 IEEE.

关键词： Robots

来源：评论

学校读者我要写书评

暂无评论

Double STAR-RIS Enhanced Secure Wireless Communications

引用

IEEE Transactions on Wireless Communications 2025年

作者： Cai, Yujin Yu, Wenwu Nie, Xiaokai Cheng, Qiang Cui, Tiejun Southeast University School of Mathematics Jiangsu Nanjing211189 China Purple Mountain Laboratories Nanjing211111 China Southeast University School of Automation Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing210096 China Southeast University Shenzhen Research Institute Shenzhen518057 China Southeast University State Key Laboratory of Millimeter Waves Nanjing210096 China

Simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RISs) have recently been employed to enhance the secure transmission due to the ability to manipulate the full-space electromagnetic signals. Prior works mainly investigate the single STAR-RIS aided secure communication system, which does not reveal the potential of the cooperative inter-STAR-RIS link. In this paper, a double STAR-RIS enhanced secure communication network involving multiple legitimate users (LUs) and eavesdroppers (Eves) is proposed. For fairness, the minimum secrecy rate maximization problem is formulated by jointly optimizing the active beamforming and passive transmission and reflection (T&R) beamforming. To deal with the highly non-convex nonlinear optimization problem, an active and passive beamforming optimization algorithm is proposed by employing the alternating optimization method. Wherein, a low-complexity sequential constraint relaxation (SCR) based algorithm is designed to solve the passive T&R beamforming optimization problem. Simulation results demonstrate that the secure performance in the double STAR-RIS assisted system outperforms that in the single RIS/double RIS/single STAR-RIS aided network. Furthermore, the proposed algorithm significantly reduces computational time (i.e., resource allocation time) compared with the traditional SCR based beamforming optimization algorithm. © 2002-2012 IEEE.

关键词： Cooperative communication

来源：评论

学校读者我要写书评

暂无评论

Probabilistic Decomposition Transformer for Time Series Forecasting

SSRN

引用

SSRN 2022年

作者： Tong, Junlong Xie, Liping Yang, Wankou Zhang, Kanjian School of Automation Southeast University Nanjing 210096 China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing 210096 China

Time series forecasting is crucial for many fields, such as disaster warning, weather prediction, and energy consumption. The Transformer-based models are considered to have revolutionized the field of sequence modeling. However, the complex temporal patterns of the time series hinder the model from mining reliable temporal dependencies. Furthermore, the autoregressive form of the Transformer introduces cumulative errors in the inference step. In this paper, we propose the probabilistic decomposition Transformer model that combines the Transformer with a conditional generative model, which provides hierarchical and interpretable probabilistic forecasts for intricate time series. The Transformer is employed to learn temporal patterns and implement primary probabilistic forecasts, while the conditional generative model is used to achieve non-autoregressive hierarchical probabilistic forecasts by introducing latent space feature representations. In addition, the conditional generative model reconstructs typical features of the series, such as seasonality and trend terms, from probability distributions in the latent space to enable complex pattern separation and provide interpretable forecasts. Extensive experiments on several datasets demonstrate the effectiveness and robustness of the proposed model, indicating that it compares favorably with the state of the art. © 2022, The Authors. All rights reserved.

关键词： Time series

来源：评论

学校读者我要写书评

暂无评论

Transformer Based Multi-modal Fusion For Place Recognition with Self-attention Mechanism

Transformer Based Multi-modal Fusion For Place Recognition w...

引用

Chinese Automation Congress (CAC)

作者： Yan Pan Xiang Yu Zeqing Chang Fang Fang Bo Zhou Key Laboratory of Measurement and Control of Complex Systems of Engineering School of Automation Southeast University Ministry of Education Nanjing P. R. China

Autonomous driving still leaves a challenging task that how to apply the complementary information captured from different sensors, i.e. cameras and LiDAR, to handle place recognition task. In this paper, a brand new pipeline was designed to tackle the retrieval problem with a multi-modal late fusion. The light-weighted convolutional blocks is applied to encode features from images, followed by converting them into pseudo point clouds. Inspired by the progress of the Transformer, the network employs a residual point transformer module to extract feature vectors for 3D points and pseudo point clouds respectively. Finally, the two corresponding local descriptors are fused to get a robust fused global descriptor, which is capable of place recognition task after end-to-end training. Experimental results collected from sequence 00, 02, 05, 06 of KITTI dataset validate that fusing information from two modalities is able to enhance performance.

关键词： Point cloud compression Training Laser radar Three-dimensional displays Streaming media Transformers Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Optimizing Bounding Box Regression in complex Backgrounds

Optimizing Bounding Box Regression in Complex Backgrounds

引用

2024 China Automation Congress, CAC 2024

作者： Chen, Jun Zhang, Haiyan Yu, Anjun Wang, Yiwei 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 Wuhan430074 China Jiangxi Ganyue Expressway Co.Ltd. Nanchang China School of Automation China University of Geosciences Wuhan430074 China

ISBN: (纸本)9798350368604

In the context of electrical power operation sites, the automated and accurate detection of whether workers are properly equipped with safety gear, such as safety clothing, helmets, and safety ropes for high-altitude operations, is crucial for ensuring the safety of personnel. Traditional object detection models often struggle to accurately identify these critical pieces of safety equipment in complex environments, due to variable object sizes, overlapping objects, and substantial background noise. To specifically address these challenges, we propose a novel loss function-Enhanced Alignment Intersection over Union (EAIoU) loss. This loss function is designed to enhance the model’s ability to discriminate between closely spaced and overlapping safety gear in real-time, significantly improving robustness against scale variations and complex backgrounds. In experiments based on the YOLOv8s architecture, our model achieved a mean Average Precision (mAP) of 92.1% on a dataset specific to electrical power operations, markedly outperforming traditional IoU loss functions. Additionally, EAIoU demonstrated improved performance on the COCO dataset, indicating its generalization capability across various complex scenarios. This research not only enhances the accuracy and real-time performance of safety equipment detection at electric power operation sites but also opens new avenues for advanced object detection technology in complex, multi-object environments. © 2024 IEEE.

关键词： Safety devices

来源：评论

学校读者我要写书评

暂无评论

Phase-Shifting Coder: Predicting Accurate Orientation in Oriented Object Detection

arXiv

引用

arXiv 2022年

作者： Yu, Yi Da, Feipeng School of Automation Southeast University Nanjing China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Southeast University Nanjing China

With the vigorous development of computer vision, oriented object detection has gradually been featured. In this paper, a novel differentiable angle coder named phase-shifting coder (PSC) is proposed to accurately predict the orientation of objects, along with a dual-frequency version (PSCD). By mapping the rotational periodicity of different cycles into the phase of different frequencies, we provide a unified framework for various periodic fuzzy problems caused by rotational symmetry in oriented object detection. Upon such a framework, common problems in oriented object detection such as boundary discontinuity and square-like problems are elegantly solved in a unified form. Visual analysis and experiments on three datasets prove the effectiveness and the potentiality of our approach. When facing scenarios requiring high-quality bounding boxes, the proposed methods are expected to give a competitive performance. The codes are publicly available at https://***/open-mmlab/mmrotate. Copyright © 2022, The Authors. All rights reserved.

关键词： Object detection

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：