检索结果-内蒙古大学图书馆

16th International Conference on Social robotics, ICSR + AI 2024

作者： Jelínek, Matouš Fischer, Kerstin Human-Robot Interaction Lab University of Southern Denmark Sønderborg Denmark

ISBN: (纸本)9789819635245

Trust is an essential aspect of human-robot interaction. For smooth and safe interactions, a calibrated level of trust – a state in which a human’s level of trust matches their robotic counterpart's capabilities – is needed. While an obvious strategy to correct possible miscalibration is transparency about the system's capabilities, previous studies are inconclusive: In some cases, the use of transparency led to an increased level of trust, in other situations, the level of trust dropped significantly. Our hypothesis is that the inconclusive results are due to the fact that what capabilities a robot is transparent about influences the effect of transparency on trust calibration. Our research therefore investigates whether transparency utterances that differ in content can either raise or dampen the perceived level of trust and competence of a robot. In an online experiment, we tested three different verbal strategies for trust calibration. Conditions differ regarding different aspects of capability the robot is transparent about: One condition uses transparency about a low level of capability, another condition uses transparency about a high level of capability. The third condition tested a robot with no additional transparency. In addition, we tested a baseline condition that used acoustic feedback only. Participants evaluated videos of four different robots, each in one of the four different conditions. We used the Multi-Dimensional Measure of Trust scale [1] to capture the perceived level of trust. Altogether 100 subjects, recruited via Prolific, participated in the experiment. Our results confirm our assumption that it matters what capabilities the robot is transparent about. Further, we confirmed that the perceived level of robots’ competence is affected by the transparency strategies used. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Nanorobots

来源：评论

学校读者我要写书评

暂无评论

robots as People Theoretical Approaches, Ethics, and Design 6

Robots as People Theoretical Approaches, Ethics, and Design

引用

6th Social robots with AI: Prospects, Risks, and Responsible Methods Robophilosophy, RP 2024

作者： Berio, Leda Blatter, Jonas Winkle, Katie Förster, Frank Fischer, Kerstin Dept. for Philosophy Ruhr University Bochum Germany Department for Information Technology Uppsala University Sweden Adaptive Systems Research Group University of Hertfordshire United Kingdom Human-Robot Interaction Lab University of Southern Denmark Denmark

来源：评论

学校读者我要写书评

暂无评论

Towards Passive Safe Reinforcement Learning: A Comparative Study on Contact-rich robotic Manipulation

arXiv

引用

arXiv 2025年

作者： Zhang, Heng Solak, Gokhan Hjorth, Sebastian Ajoudani, Arash Human-Robot Interfaces and Interaction Lab Istituto Italiano di Tecnologia Genoa Italy Università di Genova Genoa Italy

Reinforcement learning (RL) has achieved remarkable success in various robotic tasks;however, its deployment in real-world scenarios, particularly in contact-rich environments, often overlooks critical safety and stability aspects. Policies without passivity guarantees can result in system instability, posing risks to robots, their environments, and human operators. In this work, we investigate the limitations of traditional RL policies when deployed in contact-rich tasks and explore the combination of energy-based passive control with safe RL in both training and deployment to answer these challenges. Firstly, we introduce energy-based constraints in our safe RL formulation to train passivity-aware RL agents. Secondly, we add a passivity filter on the agent output for passivity-ensured control during deployment. We conduct comparative studies on a contact-rich robotic maze exploration task, evaluating the effects of learning passivity-aware policies and the importance of passivity-ensured control. The experiments demonstrate that a passivity-agnostic RL policy easily violates energy constraints in deployment, even though it achieves high task completion in training. The results show that our proposed approach guarantees control stability through passivity filtering and improves the energy efficiency through passivity-aware training. A video of real-world experiments is available as supplementary material. We also release the checkpoint model and offline data for pre-training at Hugging Face © 2025, CC BY.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

AI-Driven Relocation Tracking in Dynamic Kitchen Environments

arXiv

引用

arXiv 2025年

作者： Esfahani, Arash Nasr Hosseini, Hamed Masouleh, Mehdi Tale Kalhor, Ahmad Sajedi, Hedieh Human and Robot Interaction Lab School of Electrical and Computer Engineering University of Tehran Tehran Iran School of Mathematics Statistics and Computer Science University of Tehran Tehran Iran

As smart homes become more prevalent in daily life, the ability to understand dynamic environments is essential which is increasingly dependent on AI systems. This study focuses on developing an intelligent algorithm which can navigate a robot through a kitchen, recognizing objects, and tracking their relocation. The kitchen was chosen as the testing ground due to its dynamic nature as objects are frequently moved, rearranged and replaced. Various techniques, such as SLAM feature-based tracking and deep learning-based object detection (e.g., Faster R-CNN), are commonly used for object tracking. Additionally, methods such as optical flow analysis and 3D reconstruction have also been used to track the relocation of objects. These approaches often face challenges when it comes to problems such as lighting variations and partial occlusions, where parts of the object are hidden in some frames but visible in others. The proposed method in this study leverages the YOLOv5 architecture, initialized with pre-trained weights and subsequently fine-tuned on a custom dataset. A novel method was developed, introducing a frame-scoring algorithm which calculates a score for each object based on its location and features within all frames. This scoring approach helps to identify changes by determining the best-associated frame for each object and comparing the results in each scene, overcoming limitations seen in other methods while maintaining simplicity in design. The experimental results demonstrate an accuracy of 97.72%, a precision of 95.83% and a recall of 96.84% for this algorithm, which highlights the efficacy of the model in detecting spatial changes. © 2025, CC BY.

关键词： Optical flows

来源：评论

学校读者我要写书评

暂无评论

Bresa: Bio-inspired Reflexive Safe Reinforcement Learning for Contact-Rich robotic Tasks

arXiv

引用

arXiv 2025年

作者： Zhang, Heng Solak, Gokhan Ajoudani, Arash Human-Robot Interfaces and Interaction Lab Istituto Italiano di Tecnologia Genoa Italy Università di Genova Genoa Italy

Ensuring safety in reinforcement learning (RL)based robotic systems is a critical challenge, especially in contact-rich tasks within unstructured environments. While the state-of-the-art safe RL approaches mitigate risks through safe exploration or high-level recovery mechanisms, they often overlook low-level execution safety, where reflexive responses to potential hazards are crucial. Similarly, variable impedance control (VIC) enhances safety by adjusting the robot’s mechanical response, yet lacks a systematic way to adapt parameters, such as stiffness and damping throughout the task. In this paper, we propose Bresa, a Bio-inspired Reflexive Hierarchical Safe RL method inspired by biological reflexes. Our method decouples task learning from safety learning, incorporating a safety critic network that evaluates action risks and operates at a higher frequency than the task solver. Unlike existing recovery-based methods, our safety critic functions at a low-level control layer, allowing real-time intervention when unsafe conditions arise. The task-solving RL policy, running at a lower frequency, focuses on high-level planning (decision-making), while the safety critic ensures instantaneous safety corrections. We validate Bresa on multiple tasks including a contact-rich robotic task, demonstrating its reflexive ability to enhance safety, and adaptability in unforeseen dynamic environments. Our results show that Bresa outperforms the baseline, providing a robust and reflexive safety mechanism that bridges the gap between high-level planning and low-level execution. Real-world experiments and supplementary material are available at project website https://***/Bresa. © 2025, CC BY.

关键词： Biohazards

来源：评论

学校读者我要写书评

暂无评论

Sensor-based complete coverage path planning in dynamic environment for cleaning robot

引用

CAAI Transactions on Intelligence Technology 2018年第1期3卷 65-72页

作者： Hong Liu Jiayao Ma Weibo Huang Human Robot Interaction Lab Shenzhen Graduate School Peking University Shenzhen 518055 People's Republic of China

Using Complete Coverage Path Planning （CCPP）, a cleaning robot could visit every accessible area in the workspace. The dynamic environment requires the higher computation of the CCPP algorithm because the path needs to be replanned when the path might become invalid. In previous CCPP methods, when the neighbours of the current position are obstacles or have been visited, it is challenging for the robot to escape from the deadlocks with the least extra time cost. In this study, a novel CCPP algorithm is proposed to deal with deadlock problems in a dynamic environment. A priority template inspired by the short memory model could reduce the number of deadlocks by giving the priority of directions. Simultaneously, a global backtracking mechanism guides the robot to move to the next unvisited area quickly, taking the use of the explored global environmental information. What＇s more, the authors extend their CCPP algorithm to a multi-robot system with a market-based bidding process which could deploy the coverage time. Experiments of apartment-like scenes show that the authors＇ proposed algorithm can guarantee an efficient collision-free coverage in dynamic environments. The proposed method performs better than related approaches on coverage rate and overlap length.

关键词：路径规划清洁机器人人工智能发展现状

来源：评论

学校读者我要写书评

暂无评论

Kinematic-Sensitivity Indices for Double Gough-Stewart Platform 12

Kinematic-Sensitivity Indices for Double Gough-Stewart Platf...

引用

12th RSI International Conference on robotics and Mechatronics, ICRoM 2024

作者： Zand, Niusha Namazian, Parsa Masouleh, Mehdi Tale University of Theran Human and Robot Interaction Lab. Tehran Iran

ISBN: (纸本)9798331529734

Many indices have been introduced to optimize the architecture of a redundant robot, which is determinant in redundancy resolution of the robot. There are several ways to compute kinematic-sensitivity for dimensionally homogeneous and nonhomogeneous Jacobian matrices. The natural consequence is to use two indices that represent the kinematic-sensitivity for rotation and point-displacement, as they have been proposed by other researchers. In this paper, we compute maximum rotation and maximum displacement of end-effector of a double Gough-Stewart platform, which has a dimensionally non-homogenous Jacobian matrix, under a unit-magnitude array of actuated joint displacement. © 2024 IEEE.

关键词： End effectors

来源：评论

学校读者我要写书评

暂无评论

Generating a General Culturing Microorganism Pattern Using a Delta Parallel robot and Cam-in-Hand Calibration Method 11

Generating a General Culturing Microorganism Pattern Using a...

引用

11th RSI International Conference on robotics and Mechatronics, ICRoM 2023

作者： Pasiar, Navid Mohammadi, Arvin Ahangar, Sajjad Khomami, Navid Asadi Masouleh, Mehdi Tale University of Tehran Human and Robot Interaction Lab. Department of Mechanical Eng. Tehran Iran University of Tehran Human and Robot Interaction Lab. Tehran Iran University of Tehran Human and Robot Interaction Lab. School of Electrical and Computer Eng. Tehran Iran

ISBN: (纸本)9798350308105

robots can play a vital role in laboratory tasks, especially in culturing microorganisms. Currently, many of these operations are performed manually, which leads to biased and irreproducible results. This paper explores a calibration method for minimizing errors of a performing two certain laboratory tasks of swabbing and pipetting. This leads to enhanced accuracy and productivity. This research intends to culture microorganisms in a petri-dish in a specific pattern, measuring the coordinates of each object in the output image so the object location in the real world can be calculated. This research investigates the Cam-in-Hand method for a Delta Parallel robot, meaning that the camera is mounted on the End-Effector. In this approach, a method of translating image coordinates into real-world coordinates is introduced. It is essential to provide an additional coordinate (in this case $z -$value) before the conversion, given that the camera outputs a 2D image. A vector-based image, in SVG format, as input to the algorithm generates a set of coordinates which determines the main points for swabbing or pipetting operation. Using this data, an array of coordinates is linearly interpolated in 3D space for the swabbing operation. Conversely, the trajectory generated for pipetting uses the 4-5-6-7 interpolating polynomial. The robot then adheres to following the interpolated array of coordinates as a function of time, using a PID controller. The pipetting device is built with a 3D printer using PLA materials. The calibration is done in different heights. The Camin-Hand method leads to ±1 cm precision as a result. Minimum, maximum, and mean errors are 0.092,1.61 and 0.76 respectively for the central point. © 2023 IEEE.

关键词： Calibration

来源：评论

学校读者我要写书评

暂无评论

Multi-modal belief updates in multi-robot human-robot dialogue interactions

Multi-modal belief updates in multi-robot human-robot dialog...

引用

AISB/IACAP World Congress 2012: Linguistic and Cognitive Approaches to Dialogue Agents, Part of Alan Turing Year 2012

作者： Briggs, Gordon Scheutz, Matthias Human-Robot Interaction Lab Tufts University Medford MA 02155 United States

ISBN: (纸本)9781908187161

humans working in teams typically use task-based natural language dialogues to coordinate activities. And they use mental models of team mates which they update automatically based on perceived and communicated information to predict the actions of their team mates. It is thus reasonable to assume that humans will expect future robots interacting with humans in natural language as part of mixed-initiative teams to exhibit the same kinds of belief modeling exhibited by humans. In this paper, we propose principles that robots can use to represent beliefs and goals of other agents based on task-based natural language dialogues and use automatic inference based on communicated information to update their mental models of other agents. We demonstrate the proposed principles in a simple case study involving two robots and a human operator performing simple tasks in a laboratory environment.

关键词： Cognitive systems

来源：评论

学校读者我要写书评

暂无评论

Sound Source Localization in a 3D Multi-Surface Environment Using Learning-Based Methods 12

Sound Source Localization in a 3D Multi-Surface Environment ...

引用

12th RSI International Conference on robotics and Mechatronics, ICRoM 2024

作者： Esfahanian, Soroush Rayati, Keyhan Masouleh, Mehdi Tale University of Tehran Human and Robot Interaction Lab Electrical and Computer Engineering Tehran Iran

ISBN: (纸本)9798331529734

This study presents a three-dimensional localization of a ball drop in a multi-surface environment using cost-effective data acquisition devices, and proposes two learning-based methods to improve the baseline classical localization method based on GCC-PHAT-estimated TDOA values. The first proposed method employs a ball drop surface classification executed by Random Forest and XGBoost models to enhance localization performance, whereas the second method conducts region classification by using both basic and multi-branch neural network models, achieving improved and more efficient localization performance. A dataset of real-world sparse ball drop samples was collected in a controlled multi-surface environment for experimental execution, where results demonstrate that the two proposed methods improve the localization performance across various scenarios, with the region classification method significantly outperforming the baseline method, achieving a 17% reduction in the Mean Euclidean Distance error metric. Furthermore, the surface and region classification methods reduce the localization time compared to the baseline method by 50% and 97%, respectively. The obtained results reveal the potential of combining simple learning-based models with affordable microphone sensors for achieving an accurate localization of sparse audio samples in a multi-surface environment. © 2024 IEEE.

关键词： Deep neural networks

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：