检索结果-内蒙古大学图书馆

41st International Conference on Machine Learning, ICML 2024

作者： Yu, Le Li, Xinde Zhang, Pengfei Zhang, Zhentong Dunkin, Fir School of Automation Southeast University Nanjing China Nanjing Center for Applied Mathematics Nanjing China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Southeast University Nanjing China Southeast University Shenzhen Research Institute Shenzhen 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, MAML-based 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. Copyright 2024 by the author(s)

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Lightweight Steel Strip Defect Detection Method Based on YOLO 43

A Lightweight Steel Strip Defect Detection Method Based on Y...

引用

43rd Chinese Control Conference, CCC 2024

作者： Qian, Hao Xu, Chi China University of Geosciences School of Automation Wuhan430074 China Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Wuhan430074 China Ministry of Education Engineering Research Center of Intelligent Technology for Geo-Exploration Wuhan430074 China

ISBN: (纸本)9789887581581

As an essential component of modern industry, steel strips play an indispensable role in various fields and serve as a crucial raw material in industrial production. However, due to various factors such as production technology, the surface of steel strips often exhibits unavoidable defects like scratches and crazings. Therefore, implementing online surface quality inspection for steel strips is one effective way to enhance product quality and production efficiency. This paper proposes a detection method for surface defects in steel strips based on YOLO. By optimizing the network structure of YOLOv5s, the performance on the NEU-DET dataset has improved by nearly 4 % compared to the original network's mAP value. Experimental results indicate that the detection performance of this method is significantly superior to the current mainstream lightweight detection networks. © 2024 Technical Committee on Control Theory, Chinese Association of Automation.

关键词： Strip metal

来源：评论

学校读者我要写书评

暂无评论

Spatial-temporal structure functions in Burgers turbulence driven by an Ornstein-Uhlenbeck process

引用

Physical Review Fluids 2023年第4期8卷 044602-044602页

作者： Jin-Han Xie Department of Mechanics and Engineering Science College of Engineering State Key Laboratory for Turbulence and Complex Systems Peking University Beijing 100871 China Joint Laboratory of Marine Hydrodynamics and Ocean Engineering Laoshan Laboratory Shandong 266237 China

We explore the spatial-temporal structure functions of Burgers turbulence driven by a temporal Ornstein-Uhlenbeck (OU) process, where the characteristic time scale of the OU process is much larger than that of the energy flux across spatial scales. Based on the Kármán-Howarth-Monin equation and the temporal scale separation, we postulate an expression for the third-order spatial-temporal structure function away from the dissipation scale. This expression combines Kolmogorov's exact result of spatial structure function and the exponential temporal decay of the external force. We numerically justify this expression and find that the high-order structure functions also decay exponentially; however, the dependence of decay rates on order is different for the odd- and even-order structure functions. Comparing the OU-driven Burgers turbulence with that driven by temporal white noise, their spatial structure functions are identical when the energy injection rates are the same, which justifies Kolmogorov's theory, but these two systems' temporal structure functions differ. Also, the velocity probability density function in the OU-driven Burgers turbulence shows a bimodal distribution, contradicting the near-Gaussian distribution in white-noise-driven turbulence. Even though we lack a rigorous and general deviation of the dependence of the third-order structure function on the temporal statistics of the driving force, our results imply that one can obtain forcing temporal statistics using measured temporal data. However, inferring energy flux across spatial scales solely based on temporal information seems impossible.

关键词： Turbulence

来源：评论

学校读者我要写书评

暂无评论

Finite-Time Observer Based Integral Terminal SMC for Converter-Driven Motor systems with Mismatched Disturbances

Finite-Time Observer Based Integral Terminal SMC for Convert...

引用

2023 China Automation Congress, CAC 2023

作者： Zhang, Zhongding Guo, Zeyu Wang, Zuo Li, Shihua School of Automation Southeast University Nanjing210096 China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing210096 China Shenzhen Research Institute Southeast University Shenzhen518063 China

ISBN: (纸本)9798350303759

Converter-driven motor systems play crucial roles in contemporary industrial applications owning to the advantages of smooth starting and stepless speed regulation. For such fourth-order systems with both immeasurable states and mismatched disturbances, a finite-time control solution is proposed to address the speed tracking problem in this paper. Firstly, a finite-time observer is constructed to precisely estimate the immeasurable states and mismatched disturbances. Then, based on integral terminal sliding mode control (ITSMC), a continuous finite-time controller is designed for the systems to achieve both rapid response and superior speed tracking performance. Furthermore, a rigorous global finite-time stability analysis for the closed-loop system is presented. Finally, simulation results validate the effectiveness of the proposed control method. As compared with existing methods, the composite ITSMC algorithm design demonstrates reliable finite-time speed tracking performance and exceptional capability in disturbance rejection. © 2023 IEEE.

关键词： Disturbance rejection

来源：评论

学校读者我要写书评

暂无评论

Diving into Continual Ultra-fine-grained Visual Categorization

Diving into Continual Ultra-fine-grained Visual Categorizati...

引用

2023 International Conference on Digital Image Computing: Techniques and Applications, DICTA 2023

作者： Zhang, Pengcheng Yu, Xiaohan Bai, Xiao Zheng, Jin Wu, Xiaoyu Gao, Yongsheng Beihang University School of Computer Science and Engineering State Key Laboratory of Software Development Environment Jiangxi Research Institute Beijing China Griffith University Institute for Integrated and Intelligent Systems Nathan4111 Australia

ISBN: (纸本)9798350382204

Recent advance in ultra-fine-grained visual categorization (ultra-FGVC) has significantly boosted the capability of deep neural networks for ultra-FGVC tasks. However, building models for continually learning to recognize increasing ultra-fine-grained categories is still under-explored. This limits the application of ultra-FGVC techniques in real-world production. To this, we take the first attempt for continual ultra-FGVC. By evaluating existing continual learning methods on the constructed continual ultra-FGVC benchmark, we observe that the main bottleneck lies in the limited model plasticity for incrementally adapting to new tasks. This can be caused by excessive anti-forgetting constraints as the difficult ultra-FGVC task requires substantial update of parameters, and over-fitting on early tasks given that the ultra-fine-grained categories are with very few training samples. To tackle these problems, we propose a joint self-supervised learning and prompting model. The prompt-based continual learning framework offers proper anti-forgetting operation by fixed pretrained vision transformer and adaptive prompt selection. By jointly optimizing the learnable prompts with an adversarial self-supervised loss, the over-fitting on each continual learning task is mitigated. Extensive experiments demonstrate that the proposed method outperforms existing continual learning methods on the challenging continual ultra-FGVC problem. © 2023 IEEE.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Multi-Modal Based Fuzzy Atmosfield in Human-Robot Interaction 42

Multi-Modal Based Fuzzy Atmosfield in Human-Robot Interactio...

引用

42nd Chinese Control Conference, CCC 2023

作者： Li, Xin-Heng Liu, Zhen-Tao Wuhan430074 China Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Wuhan430074 China Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education Wuhan430074 China

ISBN: (纸本)9789887581543

Recently, the emotional robot has basic functions of perceiving and expressing emotions, but it still hard to communicate naturally between humans and robots. One major reason is that communication atmosphere is seldom considered in Human-robot Interaction (HRI). We propose Multi-modal (i.e., music background, speech, and semantics) Based Fuzzy Atmosfield (FA), which can not only realize the dynamic adjustment of FA but also dynamically regulate human emotions. In the experiment, a Pepper robot is used and one hundred volunteers are invited for HRI, and soothing piano pieces are used as background music. Questionnaires were filled by the volunteers after the experiments, from which the results show that 90% of the volunteers felt the dynamic changes in the communication atmosphere and 77% of the volunteers felt significant emotional regulation, which demonstrates the effectiveness of our method. © 2023 Technical Committee on Control Theory, Chinese Association of Automation.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

SSpose: Self-Supervised Spatial-Aware Model for Human Pose Estimation

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第11期5卷 5403-5417页

作者： Yu, Linfang Qin, Zhen Xu, Liqun Qin, Zhiguang Choo, Kim-Kwang Raymond University of Electronic Science and Technology of China Network and Data Security Key Laboratory of Sichuan Province Chengdu610054 China Industrial Research Institute Chengdu610213 China University of Electronic Science and Technology of China Network and Data Security Key Laboratory of Sichuan Province School of Information and Software Engineering Chengdu610054 China Industrial Research Institute Chengdu610213 China University of Texas at San Antonio Department of Information Systems and Cyber Security San AntonioTX78249 United States

Human pose estimation (HPE) relies on the anatomical relationships among different body parts to locate keypoints. Despite the significant progress achieved by convolutional neural networks (CNN)-based models in HPE, they typically fail to explicitly learn the global dependencies among various body parts. To overcome this limitation, we propose a spatial-aware HPE model called SSpose that explicitly captures the spatial dependencies between specific key points and different locations in an image. The proposed SSpose model adopts a hybrid CNN-Transformer encoder to simultaneously capture local features and global dependencies. To better preserve image details, a multiscale fusion module is introduced to integrate coarse- and fine-grained image information. By establishing a connection with the activation maximization (AM) principle, the final attention layer of the Transformer aggregates contributions (i.e., attention scores) from all image positions and forms the maximum position in the heatmap, thereby achieving keypoint localization in the head structure. Additionally, to address the issue of visible information leakage in convolutional reconstruction, we have devised a self-supervised training framework for the SSpose model. This framework incorporates mask autoencoder (MAE) technology into SSpose models by utilizing masked convolution and hierarchical masking strategy, thereby facilitating efficient self-supervised learning. Extensive experiments demonstrate that SSpose performs exceptionally well in the pose estimation task. On the COCO val set, it achieves an AP and AR of 77.3% and 82.1%, respectively, while on the COCO test-dev set, the AP and AR are 76.4% and 81.5%. Moreover, the model exhibits strong generalization capabilities on MPII. © 2024 IEEE.

关键词： Supervised learning

来源：评论

学校读者我要写书评

暂无评论

MUSE: A Multi-Feature Semantic Fusion Method for ROS Node Search Based on Knowledge Graph

MUSE: A Multi-Feature Semantic Fusion Method for ROS Node Se...

引用

Asia-Pacific Conference on software engineering

作者： Yuxin Zhao Xinjun Mao Tanghaoran Zhang Zhang Zhang College of Computer National University of Defense Technology Changsha China Key Laboratory of Software Engineering for Complex Systems National University of Defense Technology Changsha China

Reusing ROS components, specifically ROS Nodes, is crucial for improving the efficiency and quality of robotic software development. However, developers face challenges in finding the desired ROS Nodes for reuse due to scattered organization of ROS Nodes and the ambiguity in their ROS Node name. To address these challenges, this paper proposes a MUlti-feature SEmantic fusion method (MUSE) that leverages a domain-specific ROS knowledge graph for searching ROS Nodes. Firstly, a large dataset is constructed, comprising code files and textual descriptions related to ROS Nodes obtained from GitHub and ROS Wiki. Secondly, an in-depth analysis of user queries regarding the reuse of ROS Nodes is conducted, leading to the selection of multiple features that provide a comprehensive representation of ROS Node semantics, including Function, Hardware, Input, and Output. Subsequently, a knowledge graph of ROS Nodes is developed based on the dataset, incorporating the selected features. This knowledge graph effectively organizes scattered knowledge and resolves the issue of diverse mentions through entity disambiguation and resolution. To eliminate the semantic gap between the descriptions of features mentioned in user queries and the entities in the knowledge graph, a pretrained transformer-based model was used to measure the multi-feature semantic similarity between user queries and ROS Nodes knowledge. Finally, we employ a linear regression model to integrate the multi-feature knowledge between user queries and ROS Nodes knowledge. The proposed method has shown a 20% improvement in performance on NDCG@1 compared to other ROS Node search methods. Further evaluations highlight the effectiveness of each feature incorporated in the knowledge graph, as well as the significance of each parameter within the regression model. These findings underscore the robustness of this research in optimizing the reuse of ROS Nodes and facilitating the development of robotics software.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Bipartite consensus tracking for antagonistic topologies with leader's unknown input

Bipartite consensus tracking for antagonistic topologies wit...

引用

作者： Nan, Xiaoya Lv, Yuezu Duan, Zhisheng State Key Laboratory for Turbulence and Complex Systems Department of Mechanics and Engineering Science College of Engineering Peking University Beijing China School of Mathematics Southeast University Nanjing China

In this paper, the bipartite consensus tracking problem of linear multiagent systems on leader-follower signed directed graph is studied, where the leader's unknown input is taken into consideration. Based on only local relative information of neighboring agents, discontinuous distributed adaptive protocols are proposed, where nonlinear function is introduced for each agent to tackle leader's unknown input. To avoid tremors phenomenon, continuous distributed protocols are further presented to achieve practical bipartite consensus. Both the state feedback and output feedback cases are investigated. Different from existing literature showing bipartite consensus achievement with gauge transformation, this paper reveals a novel property of the signed Laplacian matrix, which plays a key role in simplifying the bipartite consensus demonstration. Two practical examples on synchronization of Chua's circuits are presented to illustrate the effectiveness of the designed protocols. © 2021 Chinese Automatic Control Society and John Wiley & Sons Australia, Ltd.

关键词： Multi agent systems

来源：评论

学校读者我要写书评

暂无评论

MD-TransUNet: TransUNet with Multi-attention and Dilated Convolution for Brain Stroke Lesion Segmentation 19th

MD-TransUNet: TransUNet with Multi-attention and Dilated C...

引用

19th EAI International Conference on Collaborative Computing: Networking, Applications and Worksharing, CollaborateCom 2023

作者： Xu, Jie Wan, Jian Zhang, Xin School of Computer Science and Technology Hangzhou Dianzi University Zhejiang310018 China Key Laboratory of Complex Systems Modeling and Simulation Ministry of Education Ministry of Education Beijing China School of Information and Electronic Engineering Zhejiang University of Science and Technology Zhejiang China Key Laboratory of Marine Ecosystem Dynamics Second Institute of Oceanography Ministry of Natural Resources Hangzhou China

ISBN: (纸本)9783031545276

The accurate segmentation of stroke lesion regions holds immense significance in shaping treatment strategies and rehabilitation protocols. Due to the large difference in the volume of stroke lesion areas and the great similarity between lesion areas and normal tissues, most of the existing methods for lesion segmentation cannot deal with these problems well. This paper proposes a novel network named MD-TransUNet for the segmentation of stroke lesions, whose framework is based on the UNet architecture. To fully obtain deep image features, it uses ResNet50 for downsampling. MD (multi-dilated) module is employed as the skip connection to gain more receptive fields. Different receptive fields can adapt to varying volumes of lesion areas. Then, a feature extraction module with multi-level attention mechanism is designed using ConvLSTM, non-local spatial attention, and channel attention modules to suppress useless information expression in skip connections and upsampling processes while focusing more on effective spatial and channel information in features. The experiments show that our proposed network gets superior performance than benchmark methods and indicates the generalization and effectiveness of the proposed model. © ICST Institute for Computer Sciences, Social Informatics and Telecommunications engineering 2024.

关键词： Convolution

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：