检索结果-内蒙古大学图书馆

Solving Disaster-response Tasks at a Nuclear Power Plant under Realistic Conditions

学校读者我要写书评

暂无评论

Solving Disaster-response Tasks at a Nuclear Power Plant und...

IEEE International Conference on Autonomous Robot Systems and Competitions (ICARSC)

作者： Raimund Edlinger Stefan Dumberger Christoph Fö Ls Dominik Zouhar Andreas Nü Chter Smart Automation and Robotics University of Applied Sciences Upper Austria Wels Austria Informatics VII - Robotics and Telematics Julius-Maximilians-University W&#x00FC rzburg W&#x00FC rzburg Germany

Environmental disasters usually trigger devastating damage and, in the worst case, can also cause reactor accidents in nuclear power plants, see Fukushima disaster or other detailed analysis and reports of accident’s causes 1 . The correct assessment of natural disasters and the resulting combination with radiological and nuclear accidents, is a major challenge for rescue teams. The conditions under which rescue and recovery operations, as well as exploration and transport tasks, take place are often challenging for emergency responders and involve a certain potential for danger. Serious incidents in the last decades have shown that rescue teams have an urgent need of robotic platforms for fast exploration and scene understanding, especially in scenarios involving radiation. This paper describes the rescue robotic systems used in a test scenario of a nuclear power plant in which the robots must handle sources with real radiation and lessons learned from various application-specific tasks at the EnRicH competition.

关键词： Conferences Task analysis Inductors Power generation Accidents Autonomous robots

Leveraging Semantic Graphs for Efficient and Robust LiDAR SLAM

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Wang, Neng Lu, Huimin Zheng, Zhiqiang Wang, Hesheng Liu, Yun-Hui Chen, Xieyuanli College of Intelligence Science and Technology National Key Laboratory of Equipment State Sensing and Smart Support National University of Defense Technology China IRMV Lab Department of Automation Shanghai Jiao Tong University China T Stone Robotics Institute Department of Mechanical and Automation Engineering Chinese University of Hong Kong Hong Kong

Accurate and robust simultaneous localization and mapping (SLAM) is crucial for autonomous mobile systems, typically achieved by leveraging the geometric features of the environment. Incorporating semantics provides a richer scene representation that not only enhances localization accuracy in SLAM but also enables advanced cognitive functionalities for downstream navigation and planning tasks. Existing point-wise semantic LiDAR SLAM methods often suffer from poor efficiency and generalization, making them less robust in diverse real-world scenarios. In this paper, we propose a semantic graph-enhanced SLAM framework, named SG-SLAM, which effectively leverages the geometric, semantic, and topological characteristics inherent in environmental structures. The semantic graph serves as a fundamental component that facilitates critical functionalities of SLAM, including robust relocalization during odometry failures, accurate loop closing, and semantic graph map construction. Our method employs a dual-threaded architecture, with one thread dedicated to online odometry and relocalization, while the other handles loop closure, pose graph optimization, and map update. This design enables our method to operate in real time and generate globally consistent semantic graph maps and point cloud maps. We extensively evaluate our method across the KITTI, MulRAN, and Apollo datasets, and the results demonstrate its superiority compared to state-of-the-art methods. Our method has been released at https://***/nubot-nudt/SG-SLAM. Copyright © 2025, The Authors. All rights reserved.

关键词： Graph theory

View-Robust Neural Networks for Unseen Human Action Recognition in Videos

学校读者我要写书评

暂无评论

View-Robust Neural Networks for Unseen Human Action Recognit...

2022 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2022

作者： Yu, Jiahui Ma, Tianyu Ju, Zhaojie Chen, Hang Xu, Yingke Zhejiang University Department of Biomedical Engineering Hangzhou China Binjiang Institute of Zhejiang University Innovation Center for Smart Medical Technologies & Devices Hangzhou China Shenyang Ligong University School of Automation and Electrical Engineering Shenyang China Shenyang Institute of Automation Institutes for Robotics and Intelligent Manufacturing Chinese Academy of Sciences Shenyang China University of Portsmouth School of Computing Portsmouth United Kingdom

ISBN: (数字)9781665452588

ISBN: (纸本)9781665452588

Data-driven deep learning achieved excellent performance for human action recognition. However, unseen action recognition remains a challenge for most existing neural networks. Because the action categories, collection perspectives, and scenarios considered during data collection are limited. Compared with class-unseen action recognition, view-unseen action recognition in videos is under-explored. This paper proposes view-robust neural networks (VR-Net) to recognize unseen actions in videos. The VR-Net consists of a 3D pose estimation module, skeleton adaptive transformation neural networks, and classification modules. We first extract 3D skeleton models from the video sequence based on existing pose estimation methods. Next, we propose a skeleton representation transformation scheme and achieve it based on Convolutional Neural Networks (VR-CNN) and Graph Neural Networks (VR-GCN), resulting in the optimal skeleton representations. Futhermore, we explore an associate optimization scheme and a fused output method. We evaluate the proposed neural networks on three challenging benchmarks, i.e., NTU RGB-D dataset (NTU), Kinetics-400 dataset, and Human3.6M dataset (H3.6M). The experimental results show that view robust neural networks achieve the top performance compared to state-of-the-art RGB-based and skeleton-based works, such as 93.6% on the NTU (CV) and 94.6% on the Kinetics-400 dataset (Top-5). The proposed neural networks significantly improve the recognition performance for unseen action recognition, such as 86.8% on the H3.6M (View 2). © 2022 IEEE.

关键词： Convolutional neural networks

Distribution Steering for Discrete-Time Uncertain Ensemble Systems

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Wu, Guangyu Tsiotras, Panagiotis Lindquist, Anders The Department of Automation Shanghai Jiao Tong University Shanghai China The School of Aerospace Engineering The Institute for Robotics and Intelligent Machines Georgia Institute of Technology Atlanta United States School of Artificial Intelligence Anhui University Hefei China The Smart Sensor Fusion Research Laboratory Shanghai China

Ensemble systems appear frequently in many engineering applications and, as a result, they have become an important research topic in control theory. These systems are best characterized by the evolution of their underlying state distribution. Despite the work to date, few results exist dealing with the problem of directly modifying (i.e., "steering") the distribution of an ensemble system. In addition, in most of the existing results, the distribution of the states of an ensemble of discrete-time systems is assumed to be Gaussian. However, in case the system parameters are uncertain, it is not always realistic to assume that the distribution of the system follows a Gaussian distribution, thus complicating the solution of the overall problem. In this paper, we address the general distribution steering problem for first-order discrete-time ensemble systems, where the distributions of the system parameters and the states are arbitrary with finite first few moments. Both linear and nonlinear system dynamics are considered using the method of power moments to transform the original infinite-dimensional problem into a finite-dimensional one. We also propose a control law for the ensuing moment system, which allows us to obtain the power moments of the desired control inputs. Finally, we solve the inverse problem to obtain the feasible control inputs from their corresponding power moments. We provide numerical results to validate our theoretical developments. These include cases where the parameter distribution is uniform, Gaussian, non-Gaussian, and multi-modal, respectively. © 2024, CC BY-NC-ND.

关键词： Method of moments

A surgical dataset from the da Vinci Research Kit for task automation and recognition

学校读者我要写书评

暂无评论

A surgical dataset from the da Vinci Research Kit for task a...

Electrical, Computer, Communications and Mechatronics Engineering (ICECCME), International Conference on

作者： Irene Rivas-Blanco Carlos J. Pérez Del-Pulgar Andrea Mariani Giuseppe Tortora Antonio J. Reina Dept. Systems Engineering and Automation University of Malaga Malaga Spain Dept. Excellence in Robotics & AI The BioRobotics Institute Scuola Superiore Sant’Anna Pisa Italy Smart Medical Theatre Lab ABzero srl Pisa Italy

The use of large datasets is essential in surgical robotics to advance in the field of recognition and automation of surgical tasks. Furthermore, public datasets allow the comparison of different algorithms and methods to evaluate their performance. The objective of this work is to provide a complete dataset of three common training surgical tasks performed with the da Vinci Research Kit (dVRK). The dataset contains a total of 206 trials performed by twelve different subjects. For each trial, the dataset includes 154 kinematic variables from the dVRK (both master and slave sides) together with the associated video recorded with a camera. All the data has been carefully timestamped and provided in a readable csv format. A MATLAB interface integrated with ROS for using and replicating the data is also provided.

关键词：

Image-Goal Navigation Using Refined Feature Guidance and Scene Graph Enhancement

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Feng, Zhicheng Chen, Xieyuanli Shi, Chenghao Luo, Lun Chen, Zhichao Liu, Yun-Hui Lu, Huimin The College of Intelligence Science and Technology The National Key Laboratory of Equipment State Sensing and Smart Support National University of Defense Technology China Zhejiang University China Jiangxi University of Science and Technology China The T Stone Robotics Institute Department of Mechanical and Automation Engineering The Chinese University of Hong Kong China

In this paper, we introduce a novel image-goal navigation approach, named RFSG. Our focus lies in leveraging the fine-grained connections between goals, observations, and the environment within limited image data, all the while keeping the navigation architecture simple and lightweight. To this end, we propose the spatial-channel attention mechanism, enabling the network to learn the importance of multi-dimensional features to fuse the goal and observation features. In addition, a self-distillation mechanism is incorporated to further enhance the feature representation capabilities. Given that the navigation task needs surrounding environmental information for more efficient navigation, we propose an image scene graph to establish feature associations at both the image and object levels, effectively encoding the surrounding scene information. Cross-scene performance validation was conducted on the Gibson and HM3D datasets, and the proposed method achieved state-of-the-art results among mainstream methods, with a speed of up to 53.5 frames per second on an RTX3080. This contributes to the realization of end-to-end image-goal navigation in real-world scenarios. The implementation and model of our method have been released at: https://***/nubot-nudt/RFSG. Copyright © 2025, The Authors. All rights reserved.

关键词： Image enhancement

Transitive Vision-Language Prompt Learning for Domain Generalization

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Wang, Liyuan Jin, Yan Chen, Zhen Wu, Jinlin Li, Mengke Lu, Yang Wang, Hanzi Fujian Key Laboratory of Sensing and Computing for Smart City School of Informatics Xiamen University Xiamen China Shenzhen China Centre for Artificial Intelligence and Robotics Hong Kong Institute of Science Innovation Chinese Academy of Sciences Hong Kong Institute of Automation Chinese Academy of Sciences Beijing China

The vision-language pre-training has enabled deep models to make a huge step forward in generalizing across unseen domains. The recent learning method based on the vision-language pre-training model is a great tool for domain generalization and can solve this problem to a large extent. However, there are still some issues that an advancement still suffers from trading-off between domain invariance and class separability, which are crucial in current DG problems. In this paper, we introduce a novel prompt learning strategy that leverages deep vision prompts to address domain invariance while utilizing language prompts to ensure class separability, coupled with adaptive weighting mechanisms to balance domain invariance and class separability. Extensive experiments demonstrate that deep vision prompts effectively extract domain-invariant features, significantly improving the generalization ability of deep models and achieving state-of-the-art performance on three datasets. Copyright © 2024, The Authors. All rights reserved.

关键词： Learning systems

Design and Implementation of a Robotic System for Oropharyngeal Swab Sampling Application

学校读者我要写书评

暂无评论

Design and Implementation of a Robotic System for Oropharyng...

2023 IEEE International Conference on robotics and Biomimetics, ROBIO 2023

作者： Dong, Jiaxiang Lu, Xiaowei Duan, Lihong Liu, Qi Zhang, Zhi Liu, Zhen Zhang, Xin Zhang, Xiaohua Wang, Chunbao Guo, Jing Long, Jianjun Lu, Yongtian Wei, Jianjun Li, Peng Wu, Zhengzhi Liu, Quanquan Zhao, Xuezhi South China University of Technology School of Mechanical and Automotive Engineering Guangzhou China The First Affiliated Hospital of Shenzhen University Department of Neurology Shenzhen China Guangdong University of Technology School of Automation Guangzhou China Guangdong Mk Smart Robotics CO. Ltd Zhuhai China Guangxi University of Science and Technology School of Mechanical and Automotive Engineering Liuzhou China Guilin University of Electronic Technology School of Mechanical and Electrical Engineering Guilin China Department of Rehabilitation Shenzhen Dapeng New District Nan'Ao People Hospital Shenzhen China School of Mechanical and Electrical Engineering and Automation Shenzhen China

ISBN: (纸本)9798350325706

The Oropharyngeal (OP) swab sampling is a routine, simple and effective method for the detection of respiratory infectious diseases, especially when a large-scale infectious disease outbreak occurs, such as the newly experienced COVID-19. The traditional manual OP swab sampling process is full of difficulties, such as the high demand for professional medical staff, the risk of cross-infection, and the inflexibility of sampling time, etc., making the efficient implementation of this sampling method particularly critical. Therefore, this paper proposes a novel OP swab sampling robot system to assist the OP swab sampling task. The dual-stage series-parallel hybrid manipulator component is driven by four flexible shafts with torsion properties, which guarantees the completion of sampling tasks in the narrow space of the oral cavity. Accurate sampling point location information is provided through sensor calibration and location recognition algorithms based on deep learning, and the entire sampling task is implemented using a visual-tactile fusion control system. The preliminary experiments verified the completion and feasibility of the sampling task of the OP swab sampling system, and provided a prototype for the OP swab sampling tasks. © 2023 IEEE.

关键词：

Development of a Bedside Rehabilitation Robot for Stroke Survivors

学校读者我要写书评

暂无评论

Development of a Bedside Rehabilitation Robot for Stroke Sur...

2023 IEEE International Conference on robotics and Biomimetics, ROBIO 2023

作者： Lu, Xiaowei Liu, Quanquan Duan, Lihong Dong, Jiaxiang Zhang, Zhi Liu, Zhen Zhang, Xin Zhang, Xiaohua Wang, Chunbao Guo, Jing Wei, Jianjun Long, Jianjun Wang, Yulong Li, Peng Zhao, Xuezhi Wu, Zhengzhi The First Affiliated Hospital of Shenzhen University Department of Neurology Shenzhen China Guangdong University of Technology School of Automation Guangzhou China Guangdong Mk Smart Robotics CO. Ltd Zhuhai China South China University of Technology School of Mechanical and Automotive Engineering Guangzhou China Guangxi University of Science and Technology School of Mechanical and Automotive Engineering Liuzhou China Guilin University of Electronic Technology School of Mechanical and Electrical Engineering Guilin China Shenzhen Dapeng New District Nan'Ao People Hospital Department of Rehabilitation Shenzhen China School of Mechanical and Electrical Engineering and Automation Shenzhen China

ISBN: (纸本)9798350325706

Stroke has become one of the main causes of disability and death in the elderly population. A number of stroke survivors suffers from motor dysfunction, reducing their daily quality of life. Early intervention can effectively improve rehabilitative effectiveness. Therefore, it is necessary for patients to start rehabilitation training when they are in bed. In this paper, a bedside rehabilitation robot is proposed to assist patient's lower limb rehabilitation exercise training from bedridden stage. A motor drives a double-link mechanism to achieve linear reciprocating motion for leg extension and retracting training. Two double-link mechanisms allow stroke patients to perform different training modes. A hill model of lower leg muscle was constructed to simulate the motor workload during linear reciprocating motion of patient lower limb. The preliminary experiments demonstrated the feasibility of the bedside rehabilitation robot for lower limb exercise training of stroke survivors. © 2023 IEEE.

关键词：