检索结果-内蒙古大学图书馆

arXiv 2022年

作者： Becker, Marvin Köhler, Johannes Haddadin, Sami Müller, Matthias A. The Institute of Automatic Control Leibniz University Hannover Germany The Institute for Dynamic Systems and Control ETH Zürich ZürichCH-8092 Switzerland The Munich School of Robotics and Machine Intelligence and Chair of Robotics and System Intelligence Technical University Munich Germany

Recently, many reactive trajectory planning approaches were suggested in the literature because of their inherent immediate adaption in the ever more demanding cluttered and unpredictable environments of robotic systems. However, typically those approaches are only locally reactive without considering global path planning and no guarantees for simultaneous collision avoidance and goal convergence can be given. In this paper, we study a recently developed circular field (CF)-based motion planner [1] that combines local reactive control with global trajectory generation by adapting an artificial magnetic field such that multiple trajectories around obstacles can be evaluated. In particular, we provide a mathematically rigorous analysis of this planner for static environments in the horizontal plane to ensure safe motion of the controlled robot. Contrary to existing results, the derived collision avoidance analysis covers the entire CF motion planning algorithm including attractive forces for goal convergence and is not limited to a specific choice of the rotation field, i.e., our guarantees are not limited to a specific potentially suboptimal trajectory. Our Lyapunov-type collision avoidance analysis is based on the definition of an (equivalent) two-dimensional auxiliary system, which enables us to provide tight, if and only if conditions for the case of a collision with point obstacles. Furthermore, we show how this analysis naturally extends to multiple obstacles and we specify sufficient conditions for goal convergence. Finally, we provide challenging simulation scenarios with multiple non-convex point cloud obstacles and demonstrate collision avoidance and goal convergence. © 2022, CC BY.

关键词： Motion planning

来源：评论

学校读者我要写书评

暂无评论

Tracking holistic object representations 30

Tracking holistic object representations

引用

30th British machine Vision Conference, BMVC 2019

作者： Sauer, Axel Aljalbout, Elie Haddadin, Sami Munich School of Robotics and Machine Intelligence Technical University of Munich Munich Germany

Recent advances in visual tracking are based on siamese feature extractors and template matching. For this category of trackers, latest research focuses on better feature embeddings and similarity measures. In this work, we focus on building holistic object representations for tracking. We propose a framework that is designed to be used on top of previous trackers without any need for further training of the siamese network. The framework leverages the idea of obtaining additional object templates during the tracking process. Since the number of stored templates is limited, our method only keeps the most diverse ones. We achieve this by providing a new diversity measure in the space of siamese features. The obtained representation contains information beyond the ground truth object location provided to the system. It is then useful for tracking itself but also for further tasks which require a visual understanding of objects. Strong empirical results on tracking benchmarks indicate that our method can improve the performance and robustness of the underlying trackers while barely reducing their speed. In addition, our method is able to match current state-of-the-art results, while using a simpler and older network architecture and running three times faster. © 2019. The copyright of this document resides with its authors.

关键词： Network architecture

来源：评论

学校读者我要写书评

暂无评论

Model-Based Shared Control of a Hybrid FES-Exoskeleton: An Application in Participant-Specific Robotic Rehabilitation

Model-Based Shared Control of a Hybrid FES-Exoskeleton: An A...

引用

IEEE International Conference on Rehabilitation robotics (ICORR)

作者： Hossein Kavianirad Moein Forouhar Hamid Sadeghian Satoshi Endo Sami Haddadin Sandra Hirche Chair of Information-oriented Control School of Computation Information and Technology Technical University of Munich Munich Germany Munich Institute of Robotics and Machine Intelligence Technical University of Munich Munich Germany University of Isfahan Isfahan Iran

Hybrid exoskeleton, comprising an exoskeleton interfaced with functional electrical stimulation (FES) technique, is conceptualized to complement the weakness of each other in automated neuro-rehabilitation of sensory-motor deficits. The externally actuating exoskeleton cannot directly influence neurophysiology of the patients, while FES is difficult to use in functional or goal-oriented tasks. The latter challenge is largely inherited from the fact that the dynamics of the muscular response to FES is complex, and it is highly user- and state-dependent. Due to the retardation of the muscular contraction response to the FES profile, furthermore, a commonly used model-free control scheme, such as PID control, suffers performance. The challenge in FES control is exacerbated especially in the presence of the actuation redundancy between the volitional activity of the user, powered exoskeleton, and FES-induced muscle contractions. This study therefore presents trajectory tracking performance of the hybrid exoskeleton in a novel model-based hybrid exoskeleton scheme which entices user-specific FES model-predictive control.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Involuntary Motion in Human-Robot Interaction: Effect of Interactive User Training on the Occurrence of Human Startle-Surprise Motion

Involuntary Motion in Human-Robot Interaction: Effect of Int...

引用

IEEE International Conference on intelligence and Safety for robotics (ISR)

作者： Robin Jeanne Kirschner Lisa Burr Melanie Porzenheim Henning Mayer Saeed Abdolshah Sami Haddadin Institute for Robotics and System Intelligence Munich School of Robotics and Machine Intelligence University of Munich Munich Germany Munich Center of Technology and Sociology Technical University Munich Munich Germany

In human-robot interaction (HRI) high relative velocity between robot and human at collision may lead to severe injury. Human rapid involuntary motion (IM) may therefore be a safety hazard resulting from startle or surprise (StSu). The arousal that may cause such motion can be reduced by habituation effects, indicating that user application training leads to improved safety in HRI. We suggest using an introductory hands-on user training and investigate whether it can influence the occurrence of IM. An exploratory experiment was performed with 21 participants. During the experiment a robot frontally approached the human six times. The experiment was repeated before and after a hands-on training session. We observe whether the participants show cues of StSu at the approach and compare the quantity of these cues occurring before and after hands-on robot introductory training. Our findings show promising results for the hands-on introductory training to positively effect the psychological and physical safety in HRI.

关键词： Training Conferences Human-robot interaction Psychology Hazards Collision avoidance Robots

来源：评论

学校读者我要写书评

暂无评论

Compositional Safety Verification of Infinite Networks: A Data-Driven Approach

Compositional Safety Verification of Infinite Networks: A Da...

引用

European Control Conference (ECC)

作者： Ali Aminzadeh Abdalla Swikir Sami Haddadin Abolfazl Lavaei University of Technology Iran Munich Institute of Robotics and Machine Intelligence Technical University of Munich Germany Department of Electrical and Electronic Engineering Omar Al-Mukhtar University (OMU) Albaida Libya School of Computing Newcastle University United Kingdom

ISBN: (数字)9783907144107

ISBN: (纸本)9798331540920

This paper develops a compositional framework for formal safety verification of an interconnected network comprised of a countably infinite number of discrete-time nonlinear subsystems with unknown mathematical models. Our proposed scheme involved subdividing the infinite network problem into individual subsystems, wherein the safety concept is modelled through a robust optimization program (ROP) via a notion of local-barrier certificates (L-BC). To address the difficulties associated with solving the ROP directly, primarily due to the absence of a mathematical model, we gather finite data from subsystem trajectories and leverage them to provide a scenario optimization program (SOP). We proceed with solving the resulted SOP and construct a local-barrier certificate for each unknown subsystem with a guarantee of correctness. Finally, in accordance with some small-gain conditions, we construct a global-barrier certificate (G-BC) derived from individual local certificates of subsystems, thus guaranteeing the safety of the infinite network within infinite time horizons. The practicality of our compositional findings becomes evident through a vehicle platooning scenario, characterized by a countably infinite number of vehicles with a single leader and an unlimited number of followers.

关键词： Europe Mathematical models Safety Trajectory Arrays Optimization

来源：评论

学校读者我要写书评

暂无评论

Individualized Training of Back Muscles Using Iterative Learning Control of a Compliant Balance Board

Individualized Training of Back Muscles Using Iterative Lear...

引用

IEEE International Conference on Rehabilitation robotics (ICORR)

作者： Elisabeth Jensen Reihaneh Mirjalili Kim Peper Dennis Ossadnik Fan Wu Jan Lang Matthias Martin Florian Hetfleisch Rainer Burgkart Sami Haddadin Munich Institute of Robotics and Machine Intelligence Technical University of Munich (TUM) Munich Germany Chair of Robotics and Systems Intelligence TUM Munich Germany Department of Orthopaedics and Sports Orthopaedics University Hospital Rechts der Isar TUM School of Medicine Munich Germany Chair of Non-destructive Testing TUM School of Engineering and Design Munich Germany B&W Engineering und Datensysteme GmbH Stuttgart Germany

Here we present the GyroTrainer, a bespoke mechatronic balance board system designed to trigger activation of the back muscles while the user engages in a balance-challenging game. The GyroTrainer uses admittance control coupled with an iterative learning approach so as to tailor the admittance control parameters, i.e. difficulty level, according to the user's skill. Our experimental evaluation demonstrated that an individualized admittance control stiffness could be identified for each user, which corresponds with a desired level of difficulty and increased back muscle activity. A first game implementation demonstrates the feasibility of utilizing the GyroTrainer system and the individually identified admittance control stiffness for gamification of back muscle training.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Risk Estimation for ICU Patients with Personalized Anomaly-Encoded Bedside Patient Data

Risk Estimation for ICU Patients with Personalized Anomaly-E...

引用

Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

作者： Kai Wu Ee Heng Chen Felix Wirth Keti Vitanova Rüdiger Lange Darius Burschka German Heart Center Munich Munich Germany Department of Computer Engineering Machine Vision and Perception Group TUM School of Computation Information and Technology Technical University of Munich Garching Germany MIRMI - Munich Institute of Robotics and Machine Intelligence Technical University of Munich Munich

We propose a novel framework to estimate intensive care unit patients' health risk continuously with anomaly-encoded patient data. This framework consists of two modules. In the first module, we use Gaussian process models to learn change trend and day-night circulation in temporal patient data and annotate abnormal data. Such models provide dynamically adaptable bedside patient monitoring instead of conventional threshold-based monitoring. In the second module, we use the abnormal data together with the learned Gaussian models to estimate patients' risk level by predicting their in-hospital mortality and remaining length of stay in ICU ward. We show that prediction models with anomaly-encoded data have better performance than those with raw patient measurements, and they are comparable with state-of-art prediction models.

关键词：

来源：评论

学校读者我要写书评

暂无评论

HDCircuit: Brain-Inspired HyperDimensional Computing for Circuit Recognition

HDCircuit: Brain-Inspired HyperDimensional Computing for Cir...

引用

Design, Automation and Test in Europe Conference and Exhibition

作者： Paul R. Genssler Lilas Alrahis Ozgur Sinanoglu Hussam Amrouch Semiconductor Test and Reliability University of Stuttgart Stuttgart Germany New York University Abu Dhabi UAE Chair of AI Processor Design Technical University of Munich TUM School of Computation Information and Technology Munich Institute of Robotics and Machine Intelligence Munich Germany

ISBN: (数字)9783981926385

ISBN: (纸本)9798350348606

Circuits possess a non-Euclidean representation, necessitating the encoding of their data structure (e.g., gate-level netlists) into fixed formats like vectors. This work is the first to propose brain-inspired hyperdimensional computing (HDC) for optimized circuit encoding. HDC does not require extensive training to encode a gate-level netlist into a hypervector and simplifies the similarity check between circuits from graph-based to the similarity between their hypervectors. We introduce a versatile HDC-based encoding method for circuit encoding. We demonstrate its effectiveness with the application of circuit recognition using ITC-99 and ISCAS-85 benchmarks. We maintain a 98.2% accuracy, even when the designs are obfuscated using logic locking.

关键词： Training Circuits Intellectual property Logic gates Benchmark testing Data structures Encoding

来源：评论

学校读者我要写书评

暂无评论

Modeling Action Spatiotemporal Relationships Using Graph-Based Class-Level Attention Network for Long-Term Action Detection

Modeling Action Spatiotemporal Relationships Using Graph-Bas...

引用

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

作者： Yuankai Wu Xin Su Driton Salihu Hao Xing Marsil Zakour Constantin Patsch Chair of Media Technology and Munich Institute of Robotics and Machine Intelligence School of Computation Information and Technology Technical University of Munich Germany Maschine Vision and Perception group School of Computation Information and Technology Technical University of Munich Germany

In recent years, Action Detection has become an active research topic in various fields such as human-robot interaction and assistive robots. Most of the previous methods in this field focus on temporally processing the action representation, without considering the dependencies among the action classes. However, actions that occur in a video are constantly related, and this correlation could offer effective clues for detection tasks. In this work, we propose to exploit the information of related action classes with the help of a graph neural network in conjunction with temporal modeling. We introduce the attention-based temporal class module (ATC), which models the inherent action dependencies on the graph and learns action-specific features among temporal dimensions with a dual-branch attention mechanism. Further, we present the Graph-based Class-level Attention Network (GCAN), which is built upon ATC modules with increasing temporal receptive fields to handle actions instances in complex untrimmed videos. Our network is evaluated on two challenging benchmark datasets with dense annotations: Charades and MultiTHUMOS. Experimental results show that our approach demonstrates highly competitive results with a significantly reduced model complexity.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Continuous Risk Estimation of Acute Kidney Failure with Dense Temporal Data for ICU Patients

Continuous Risk Estimation of Acute Kidney Failure with Dens...

引用

Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

Acute kidney failure is a dangerous complication for ICU patients, and it is difficult to identify at early stage with conventional medical analysis. In recent years, machine learning approaches have been applied to tackle medical diagnosis tasks with great performance. In this work, we deploy machine learning models for early detection of acute kidney failure that can handle static, temporal, sparse and dense data of ICU patients. We investigate different pre-processing methods for patient data to achieve higher prediction performance and how they influence the contribution of different physiological signals in the prediction process.

关键词：

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：