With the increasing prevalence of complex vision-based sensing methods for use in obstacle identification and state estimation, characterizing environment-dependent measurement errors has become a difficult and essential part of modern robotics. This paper presents a self-supervised learning approach to safety-critical control. In particular, the uncertainty associated with stereo vision is estimated, and adapted online to new visual environments, wherein this estimate is leveraged in a safety-critical controller in a robust fashion. To this end, we propose an algorithm that exploits the structure of stereo-vision to learn an uncertainty estimate without the need for ground-truth data. We then robustify existing Control Barrier Function-based controllers to provide safety in the presence of this uncertainty estimate. We demonstrate the efficacy of our method on a quadrupedal robot in a variety of environments. When not using our method safety is violated. With offline training alone we observe the robot is safe, but overly-conservative. With our online method the quadruped remains safe and conservatism is reduced.
Updating Robot Safety Representations Online from Natural Language Feedback
Federal University of Minas Gerais（米纳斯吉拉斯联邦大学）；University of Rochester（罗切斯特大学）；Delft University of Technology（代尔夫特理工大学）；University of Southern California（南加利福尼亚大学）；Carnegie Mellon University（卡内基梅隆大学）
Barrier Bayesian Linear Regression: Online Learning of Control Barrier Conditions for Safety-Critical Control of Uncertain Systems
University of Toronto Institute for Aerospace Studies（多伦多大学航空航天研究所）；Technical University of Munich（慕尼黑工业大学）；University of Toronto Robotics Institute（多伦多大学机器人研究所）；Vector Institute for Artificial Intelligence（矢量人工智能研究所）
StereoVoxelNet: Real-Time Obstacle Detection Based on Occupancy Voxels from a Stereo Camera Using Deep Neural Networks
Institute for Experiential Robotics, Northeastern University（东北大学体验机器人研究所）；Khoury College of Computer Sciences, Northeastern University（东北大学胡里计算机学院）；Department of Electrical and Computer Engineering, Northeastern University（东北大学电气与计算机工程系）
障碍物检测是机器人导航中的一个安全关键问题，其中立体匹配是一种流行的基于视觉的方法。虽然深度神经网络在计算机视觉中已经显示出令人印象深刻的结果，但大多数先前的障碍物检测工作仅利用传统的立体匹配技术来满足实时的【形】反馈的计算约束。本文提出了一种计算效率高的方法，该方法采用深度神经网络直接从立体图像中检测占用。我们的方法不是从立体数据中学习点云对应关系，而是基于体积表示提取紧凑的障碍物分布。此外，我们修剪的安全无关空间的计算，在一个由粗到细的方式的基础上产生的解码器的八叉树。因此，我们在机载计算机（NVIDIA Jetson TX 2）上实现了实时的【形】性能。我们的方法可以在32米范围内准确检测障碍物，并实现更好的IoU（Intersection over Union）和CD（Chamfer Distance）分数，而计算成本仅为最先进的立体模型的2%。此外，我们验证了我们的方法的鲁棒性和现实世界的可行性，通过自主导航实验与真正的机器人。因此，我们的工作有助于缩小机器人感知中基于立体的系统与计算机视觉中最先进的立体模型之间的差距。为了应对高质量的真实室内立体数据集的稀缺性，我们使用移动机器人收集了1.36小时的立体数据集，用于微调我们的模型。有关数据集、代码和更多详细信息（包括其他可视化），请访问https://lhy.xyz/stereovoxelnet
Parameter-Robust MPPI for Safe Online Learning of Unknown Parameters
Division of Robotics, Perception and Learning, KTH Royal Institute of Technology（机器人、感知与学习系，皇家理工学院）；Digital Futures（数字未来）；Reliable Autonomous Systems Lab, Massachusetts Institute of Technology（可靠自主系统实验室，麻省理工学院）；Automatic Control Laboratory, Swiss Federal Institute of Technology in Zürich（自动控制实验室，瑞士联邦理工学院（苏黎世））
LST-SLAM: A Stereo Thermal SLAM System for Kilometer-Scale Dynamic Environments
PEAK-Lab, The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州）峰值实验室）；School of Electrical and Electronic Engineering, Nanyang Technological University（南洋理工大学电子与电气工程学院）
