检索结果-内蒙古大学图书馆

Chinese Control Conference (CCC)

作者： De-Xiu Wang Qing-Hao Meng Rui-Xue Han Hong-Yue Li Hui-Rang Hou Tianjin Key Laboratory of Process Measurement and Control Institute of Robotics and Autonomous Systems School of Electrical and Information Engineering Tianjin University Tianjin China Tianjin Navigation Instruments Research Institute Tianjin China

Identification of the type of combustion-supporting agents (CSAs) by an electronic nose (e-nose) is severely limited due to the absence of untested gas concentration in the e-nose training set. In order to solve this problem, this study designs an e-nose for classifying four CSAs (Diesel, Gasoline, Ethanol and Thinner), and builds cross-concentrations data set. This study uses Gramian Angular Field method that converts one-dimensional time series data to two-dimensional data, and proposes an algorithm named the Multi-Branch Feature Fusion Network (M-BFFNet). This algorithm passes the one-dimensional and two-dimensional data through two different networks and obtains two types of features. First, the algorithm can strength the one-dimensional raw data, integrate the advantages of different networks, and fuse the characteristics of different networks. Second, it can learn the characteristics of volatile gas with specific concentrations to identify the unknown concentrations. Last, a feature fusion module is designed to improve the algorithm ability of learning odor fingerprint. In order to verify the superiority of the algorithm, this study tests the recognition performance of the e-nose based systems and compares the proposed M-BFFNet with traditional algorithms. The results show that the proposed M-BFFNet improves the identification accuracy compared to other algorithms. According to the analysis of this work, the M-BFFNet can be used to identify CSAs with unknown concentrations.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Cross Data Set Generalization of Ultrasound Image Augmentation using Representation Learning: A Case Study

引用

Current Directions in Biomedical engineering 2021年第2期7卷 755-758页

作者： Wulff, Daniel Mehdi, Mohamad Ernst, Floris Hagenah, Jannis Institute for Robotics and Cognitive Systems University of Luebeck Ratzeburger Allee 160 Luebeck 23562 Germany Computational Health Informatics Lab Institute of Biomedical Engineering University of Oxford Old Road Campus Research Building Oxford OX37DQ United Kingdom

Data augmentation is a common method to make deep learning assessible on limited data sets. However, classical image augmentation methods result in highly unrealistic images on ultrasound data. Another approach is to utilize learning-based augmentation methods, e.g. based on variational autoencoders or generative adversarial networks. However, a large amount of data is necessary to train these models, which is typically not available in scenarios where data augmentation is needed. One solution for this problem could be a transfer of augmentation models between different medical imaging data sets. In this work, we present a qualitative study of the cross data set generalization performance of different learning-based augmentation methods for ultrasound image data. We could show that knowledge transfer is possible in ultrasound image augmentation and that the augmentation partially results in semantically meaningful transfers of structures, e.g. vessels, across domains. © 2021 by Walter de Gruyter Berlin/Boston.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Orientation Control with Variable Stiffness Dynamical systems

Orientation Control with Variable Stiffness Dynamical System...

引用

IEEE/RSJ International Conference on Intelligent Robots and systems (IROS)

作者： Youssef Michel Matteo Saveriano Fares J. Abu-Dakka Dongheui Lee Chair of Human-Centered Assistive Robotics Technical University of Munich Munich Germany Department of Industrial Engineering University of Trento Trento Italy Munich Institute of Robotics and Machine Intelligence Technical University of Munich Munich Germany Autonomous Systems TU Wien Vienna Austria German Aerospace Center (DLR) Institute of Robotics and Mechatronics Munich Germany

Recently, several approaches have attempted to combine motion generation and control in one loop to equip robots with reactive behaviors, that cannot be achieved with traditional time-indexed tracking controllers. These approaches however mainly focused on positions, neglecting the orientation part which can be crucial to many tasks e.g. screwing. In this work, we propose a control algorithm that adapts the robot's rotational motion and impedance in a closed-loop manner. Given a first-order Dynamical System representing an orientation motion plan and a desired rotational stiffness profile, our approach enables the robot to follow the reference motion with an interactive behavior specified by the desired stiffness, while always being aware of the current orientation, represented as a Unit Quaternion (UQ). We rely on the Lie algebra to formulate our algorithm, since unlike positions, UQ feature constraints that should be respected in the devised controller. We validate our proposed approach in multiple robot experiments, showcasing the ability of our controller to follow complex orientation profiles, react safely to perturbations, and fulfill physical interaction tasks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Kinematic Parameter Calibration for Parallel Robot Based on Robust Estimation and PCR Algorithm

Kinematic Parameter Calibration for Parallel Robot Based on ...

引用

Advanced Algorithms and Control engineering (ICAACE), International Conference on

作者： Jin Wang Renge Zhang Weijun Hu Haiyun Zhang Guodong Lu Silu Chen Jingbo Luo State Key Laboratory of Fluid Power and Mechatronic Systems School of Mechanical Engineering Zhejiang University Hangzhou China Robotics Institute of Zhejiang University Hangzhou China School of Robotics Ningbo University of Technology Ningbo China Zhejiang Qiantang Robot and Intelligent Equipment Research Co. LTD Hangzhou China Ningbo Institute of Materials Technology and Engineering Chinese Academy of Sciences Ningbo China

ISBN: (数字)9798350361445

ISBN: (纸本)9798350361452

Due to the existence of redundant identification parameters and parallel branch chains, the nonlinear error model of parallel robot is likely to produce multicollinearity, which will lead to the undesirable ill-conditioned identification matrix. Though PCR method can modify the ill-conditioned identification matrix, it is still susceptible to the gross error of the original data. In terms of these problems, this paper proposes a parameter calibration algorithm for parallel robot based on robust estimation and principal component regression (RE-PCR). Firstly, the linear identification model is established based on the closed-loop vector chain perturbation theory to improve the identification accuracy. On this basis, for the configuration types with redundant identification parameters and parallel branch chains, PCR is used to modify the ill-conditioned matrix and a robust estimation (RE) algorithm is designed to improve the resistance capacities of PCR to gross error data points. Finally, through the identification experiment based on 4PPa-2PaR parallel robot, the proposed algorithm demonstrates higher efficiency and stronger robustness in comparison with traditional methods.

关键词： Resistance Parallel robots Parameter estimation Perturbation methods Estimation Kinematics Vectors

来源：评论

学校读者我要写书评

暂无评论

Harvesting Curvatures for Communication-Efficient Distributed Optimization

Harvesting Curvatures for Communication-Efficient Distribute...

引用

Asilomar Conference on Signals, systems & Computers

作者： Diogo Cardoso Boyue Li Yuejie Chi João Xavier Department of Electrical and Computer Engineering Carnegie Mellon University USA Institute for Systems and Robotics IST-University of Lisbon Portugal

This paper considers the design of distributed optimization algorithms in the server-client setting, with the goal of achieving better communication efficiency, a crucial desiderata in modern distributed machine learning applications. One popular approach to reduce communication is to impose more computation between communication rounds, by letting each client optimize an approximation of the global objective function-constructed using the local objective function and aggregated history information-and exchange the updates with the parameter server for global consensus. In particular, judiciously incorporating second-order information in the local approximation often leads to better communication efficiency when the problem is ill-conditioned. However, existing methods construct the second-order correction term in an isotropic manner, without taking into account the curvature information of the global objective function accumulated over the course of the algorithm. This paper proposes a novel algorithm that refines this idea by constructing a second-order correction term using a BFGS-style update formula, where the kernel matrix is updated recursively using only history gradients to harvest curvature information for accelerating convergence. Numerical experiments demonstrate improved communication efficiency over competitive baseline algorithms on both synthetic and real datasets.

关键词： Machine learning algorithms Memory management Machine learning Approximation algorithms Linear programming Servers History

来源：评论

学校读者我要写书评

暂无评论

HG-CBS Planner: Heuristic Group-based Motion Planning for Multi-robot

HG-CBS Planner: Heuristic Group-based Motion Planning for Mu...

引用

IEEE Conference on Industrial Electronics and Applications (ICIEA)

作者： Weilong Liu Jin Wang Kewen Zhang Huan Yu Zhi Zheng Guodong Lu State Key Laboratory of Fluid Power and Mechatronic Systems School of Mechanical Engineering Zhejiang University Hangzhou China Robotics Research Center of Yuyao City Ningbo China Robotics Institute of Zhejiang University Hangzhou China College of Mechanical Engineering Zhejiang University of Technology Hangzhou China

Multi-Agent Path Finding (MAPF) has been widely studied in the field of artificial intelligence and multi-agent path search, and researchers have made improvements to make such algorithms applicable to actual nonholonomic constrained robots. In this paper, the collision detection is improved by body-occupied collision detection, which improves the efficiency of collision detection. In addition, to solve the problem of low efficiency in centralized search for multi-robot solutions, reasonable grouping is performed to reduce the number of collision nodes expanded in the centralized Conflict-based search(CBS), and feasible results are obtained faster. At the same time, the heuristic multi-robot search based on collision frequency is improved to reduce the search time. Finally, we constructed the trajectory smoothing problem based on the multi-robot path search results and applied the solution to real robot trajectory tracking. Our research provides valuable insights into solving trajectory planning problems for nonholonomic constrained mobile robots in practical scenarios.

关键词：

来源：评论

学校读者我要写书评

暂无评论

QuerySOD: A Small Object Detection Algorithm Based on Sparse Convolutional Network and Query Mechanism

QuerySOD: A Small Object Detection Algorithm Based on Sparse...

引用

IEEE/RSJ International Conference on Intelligent Robots and systems (IROS)

作者： Zhengcai Cao Junnian Li Jie Niu MengChu Zhou State Key Laboratory of Robotics and Systems Harbin Institute of Technology Harbin China College of Information Science and Technology Beijing University of Chemical Technology Beijing China Department of Electrical and Computer Engineering New Jersey Institute of Technology Newark NJ USA

ISBN: (数字)9798350377705

ISBN: (纸本)9798350377712

Although remarkable advances have been achieved in generic object detection, small object detection (SOD) remains challenging owing to small objects’ information loss and noisy representation caused by their non-uniform distribution. Their limited width and height, scale variations, and redundant computation make SOD hard. To overcome them, this work proposes a new SOD method based on sparse convolutional network (SCNet) and Query Mechanism called QuerySOD. First, an extended feature pyramid network is constructed for extracting feature maps of small objects with more regional details. Then, a Sparse Head is neatly designed by using SCNet for accelerating the interfering speed and obtaining weights of each layer. After that, a Query Mechanism is innovatively introduced for harvesting the benefit of sparse value feature maps from the Sparse Head. QuerySOD is evaluated on public benchmarks including COCO and VisDrone. Finally, we apply it on ‘Jinghai’ unmanned survey vehicles and receive excellent SOD performance from this real-world application.

关键词： Surveys Accuracy Object detection Seaports Lakes Feature extraction Prediction algorithms Convolutional neural networks Security Noise measurement

来源：评论

学校读者我要写书评

暂无评论

Strength in Diversity: Multi-Branch Representation Learning for Vehicle Re-Identification

arXiv

引用

arXiv 2023年

作者： Almeida, Eurico Silva, Bruno Batista, Jorge The Institute of Systems and Robotics University of Coimbra Coimbra3030-219 Portugal The Department of Electrical and Computers Engineering Faculty of Science and Technology University of Coimbra Coimbra3030-290 Portugal

This paper presents an efficient and lightweight multi-branch deep architecture to improve vehicle re-identification (V-ReID). While most V-ReID work uses a combination of complex multi-branch architectures to extract robust and diversified embeddings towards re-identification, we advocate that simple and lightweight architectures can be designed to fulfill the Re-ID task without compromising performance. We propose a combination of Grouped-convolution and Loss-Branch-Split strategies to design a multi-branch architecture that improve feature diversity and feature discriminability. We combine a ResNet50 global branch architecture with a BotNet self-attention branch architecture, both designed within a Loss-Branch-Split (LBS) strategy. We argue that specialized loss-branch-splitting helps to improve re-identification tasks by generating specialized re-identification features. A lightweight solution using grouped convolution is also proposed to mimic the learning of loss-splitting into multiple embeddings while significantly reducing the model size. In addition, we designed an improved solution to leverage additional metadata, such as camera ID and pose information, that uses 97% less parameters, further improving re-identification performance. In comparison to state-of-the-art (SoTA) methods, our approach outperforms competing solutions in Veri-776 by achieving 85.6% mAP and 97.7% CMC1 and obtains competitive results in Veri-Wild with 88.1% mAP and 96.3% CMC1. Overall, our work provides important insights into improving vehicle re-identification and presents a strong basis for other retrieval tasks. Our code is available at the link. Copyright © 2023, The Authors. All rights reserved.

关键词： Vehicles

来源：评论

学校读者我要写书评

暂无评论

A Context-Enhanced Full-Resolution Floor Plan Segmentation Network for Topological Semantic Mapping

A Context-Enhanced Full-Resolution Floor Plan Segmentation N...

引用

IEEE/RSJ International Conference on Intelligent Robots and systems (IROS)

作者： Zhengcai Cao Yiyang Sun Zhe Ma MengChu Zhou State Key Laboratory of Robotics and Systems Harbin Institute of Technology Harbin China College of Information Science and Technology Beijing University of Chemical Technology Beijing China Department of Electrical and Computer Engineering New Jersey Institute of Technology Newark NJ USA

ISBN: (数字)9798350377705

ISBN: (纸本)9798350377712

Topological semantic maps provide a practical solution to enhance indoor navigation for the Partially Sighted or Visually Impaired (PSVI). Segmenting indoor floor plans and extracting boundaries are key to constructing these maps. The existing methods exhibit low accuracy in segmentation. To achieve desired high segmentation accuracy, we introduce a Context-Enhanced Full-Resolution Network (CEFRN) for floor plan segmentation. It is designed to harness the shallow detailed features and inter-category contextual dependencies inherent in floor plans. CEFRN integrates modified residual blocks to capture the low-stage full-resolution features while maintaining its compactness. A position attention module is employed to refine the deep-stage contextual information. We also propose a two-dimensional deep supervision method to merge features from both stages, which significantly boosts the feature representation ability of CEFRN. Finally, a practical topological semantic mapping method for PSVI indoor navigation is introduced. Experimental results demonstrate that CEFRN’s segmentation accuracy well exceeds the state-of-the-art methods’. It can be used to well support accurate topological semantic mapping.

关键词： Accuracy Hospitals Indoor navigation Semantics Feature extraction Mobile robots Floors Intelligent robots Autonomous vehicles Drones

来源：评论

学校读者我要写书评

暂无评论

New Type of Insoles Based on Air Pressure Arrays for Fall Detection in the Elderly

New Type of Insoles Based on Air Pressure Arrays for Fall De...

引用

International Conference on Applied Nonlinear Dynamics, Vibration and Control, ICANDVC 2024

作者： Xu, Yiwei Lu, Junjie Wan, Xiao Chang, Tao Lei, Qian Huang, Zixin School of Electrical and Information Engineering Wuhan Institute of Technology Wuhan430205 China Hubei Key Laboratory of Digital Textile Equipment Wuhan Textile University Wuhan430200 China Institute of Robotics and Automatic Information Systems College of Artificial Intelligence Tianjin300350 China

ISBN: (纸本)9789819633166

For traditional fall detection insoles, the number of sensors is large, the cost is high, the sensor must correspond to the detection site one by one, and the resolution is low when the distance is far from the geometric center of the sensor, and the high sensitivity cannot be maintained in a large range. The rigid conduction of the force formed by the sensor and the soles of the elderly leads to poor comfort and low trust of the machine. A new soft insole for fall detection, based on an air pressure array, was designed and produced. Insole uses BMP280 pressure sensor and a 3D printed mold for silicone multi-layer injection molding. With four built-in airbags in one foot, tactile pressure can be converted to air pressure. STM32 microprocessor is used to construct the pressure information acquisition and processing system. Finally, Kalman filter algorithm is used to process the collected information. The experiment shows that the pressure detection system based on pressure change has high sensitivity and small repeatability error, and can effectively detect the pressure change and distribution of the sole. And reduce the noise interference, effectively improve the comfort and improve the precision and stability of recognition and acquisition equipment at the same time reduce the cost. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Pressure sensors

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：