☆ 4.7 Article

Adaptive Robotic Information Gathering via non-stationary Gaussian processes

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2023)

Related references

Note: Only part of the references are listed.

Article Robotics

Partially Observable Markov Decision Processes in Robotics: A Survey

Mikko Lauri et al.

Summary: This article introduces the application of partially observable Markov decision process (POMDP) in robot tasks, analyzes the characteristics of these tasks, and explores the mathematical and algorithmic properties of the POMDP framework.

IEEE TRANSACTIONS ON ROBOTICS (2023)

Add to Collection

Article Robotics

Resilient Active Information Acquisition With Teams of Robots

Brent Schlotfeldt et al.

Summary: This article introduces a robust and adaptive multirobot planning algorithm RAIN, which can plan information acquisition tasks in adversarial environments and exhibits superior performance in multiple information acquisition scenarios.

IEEE TRANSACTIONS ON ROBOTICS (2022)

Add to Collection

Review Robotics

Robot Learning From Randomized Simulations: A Review

Fabio Muratore et al.

Summary: The rise of deep learning has led to a change in the paradigm of robotics research, with a preference for methods that require large amounts of data. However, it is expensive to generate such data sets on a physical platform, so state-of-the-art approaches learn in simulation and then transfer the knowledge to real robots. Simulators, although increasingly realistic, are based on models and therefore inherently imperfect. This raises the question of how simulators can be modified to enhance learning robot control policies and overcome the reality gap. Domain randomization, a technique for learning from randomized simulations, is comprehensively reviewed in this paper.

FRONTIERS IN ROBOTICS AND AI (2022)

Add to Collection

Article Robotics

Safe Active Dynamics Learning and Control: A Sequential Exploration-Exploitation Framework

Thomas Lew et al.

Summary: This article introduces a method using Bayesian meta-learning and last-layer adaptation to maintain safety in the presence of dynamic uncertainty. By combining online last-layer adaptation with tight confidence sets, safe trajectory planning is achieved to ensure system safety.

IEEE TRANSACTIONS ON ROBOTICS (2022)

Add to Collection

Article Robotics

Informative Path Planning to Estimate Quantiles for Environmental Analysis

Isabel M. Rayas Fernandez et al.

Summary: In this study, we propose a method that uses a robot for informative path planning survey to select suitable locations for scientific analysis. By developing a robot planner and novel objective functions, we are able to improve the estimation of quantile values and find corresponding locations. In the experiment, our method demonstrates higher accuracy and lower error rates.

IEEE ROBOTICS AND AUTOMATION LETTERS (2022)

Add to Collection

Proceedings Paper Automation & Control Systems

Adaptive Informative Path Planning Using Deep Reinforcement Learning for UAV-based Active Sensing

Julius Rueckin et al.

Summary: Aerial robots are being used more often for environmental monitoring and exploration, but efficiently planning paths to maximize the value of collected data in unknown environments is a challenge. To tackle this problem, the authors propose a new approach for informative path planning using deep reinforcement learning. By combining tree search with an offline-learned neural network, their method is able to handle high-dimensional state and large action spaces. Simulation results show that the approach performs as well as existing methods while reducing runtime by 8-10x.

2022 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA 2022) (2022)

Add to Collection

Proceedings Paper Automation & Control Systems

TIGRIS: An Informed Sampling-based Algorithm for Informative Path Planning

Brady Moon et al.

Summary: This article introduces a sampling-based approach to address the problem of informative path planning in high-dimensional spaces and non-trivial sensor constraints. The method utilizes informed sampling and considers potential information gain to generate an optimal path and identify objects of interest in large search spaces. It outperforms the baseline method by 18.0% in experiments.

2022 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS) (2022)

Add to Collection

Article Robotics

HyP-DESPOT: A hybrid parallel algorithm for online planning under uncertainty

Panpan Cai et al.

Summary: The Hybrid Parallel DESPOT (HyP-DESPOT) algorithm leverages both CPU and GPU parallelization to achieve real-time online planning performance for robots in uncertain, dynamic environments. It combines parallel DESPOT tree search on multi-core CPUs with parallel Monte Carlo simulations on GPUs to speed up online planning by several hundred times in challenging robotic tasks.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2021)

Add to Collection

Article Robotics

Underwater Soft Robot Modeling and Control With Differentiable Simulation

Tao Du et al.

Summary: This method utilizes differentiable simulation and analytical hydrodynamic model for modeling and controlling underwater soft robots. By alternating between simulation and experiments with real data, it effectively reduces the simulation-to-reality gap and improves the performance of controllers in real experiments.

IEEE ROBOTICS AND AUTOMATION LETTERS (2021)

Add to Collection

Article Robotics

Roadmap Learning for Probabilistic Occupancy Maps With Topology-Informed Growing Neural Gas

Manish Saroya et al.

Summary: This study introduces a topology-informed growing neural gas algorithm for generating navigation roadmaps in uncertain and cluttered environments. Experimental results demonstrate that the method outperforms baseline methods in terms of graph connectivity, path solution quality, and search efficiency, achieving similar performance to a denser PRM* with a 27x faster query time for shortest-path searches.

IEEE ROBOTICS AND AUTOMATION LETTERS (2021)

Add to Collection

Article Robotics

CRMI: Confidence-Rich Mutual Information for Information-Theoretic Mapping

Yang Xu et al.

Summary: This letter introduces information-theoretic active mapping and exploration using confidence-rich mutual information (CRMI) to improve accuracy and efficiency. By explicitly modeling measurement dependencies, a more effective robot exploration behavior is achieved.

IEEE ROBOTICS AND AUTOMATION LETTERS (2021)

Add to Collection

Article Robotics

Informed Sampling Exploration Path Planner for 3D Reconstruction of Large Scenes

Yves Kompis et al.

Summary: The study proposes an online path-planning algorithm for fast exploration and 3D reconstruction of unknown areas, utilizing informed sampling techniques to achieve higher efficiency. Results show that the method outperforms existing exploration path planners in terms of speed and reconstruction quality.

IEEE ROBOTICS AND AUTOMATION LETTERS (2021)

Add to Collection

Review Chemistry, Analytical

Active Mapping and Robot Exploration: A Survey

Iker Lluvia et al.

Summary: Simultaneous localization and mapping (SLAM) aims to build a map of the environment without prior information, using data obtained from sensors. Local SLAM allows robots to autonomously navigate while exploring the environment and constructing a map, minimizing mapping errors.

SENSORS (2021)

Add to Collection

Article Robotics

Data-Efficient Domain Randomization With Bayesian Optimization

Fabio Muratore et al.

Summary: BayRn addresses the reality gap in robot control learning by adapting domain parameter distribution through Bayesian optimization, leading to more efficient policy learning and reducing the reliance on prior knowledge.

IEEE ROBOTICS AND AUTOMATION LETTERS (2021)

Add to Collection

Article Robotics

SACBP: Belief space planning for continuous-time dynamical systems via stochastic sequential action control

Haruki Nishimura et al.

Summary: A novel belief space planning technique is proposed by considering the belief system as a hybrid dynamical system with time-driven switching, based on perturbation theory of differential equations and extending sequential action control to stochastic dynamics. The resulting algorithm, SACBP, synthesizes control signals in near real-time without discretization of spaces or time and can handle general parametric Bayesian filters. The approach demonstrates effectiveness in active sensing scenarios and model-based Bayesian reinforcement learning problems, outperforming existing solution techniques like approximate dynamic programming and local trajectory optimization.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2021)

Add to Collection

Review Automation & Control Systems

Active learning in robotics: A review of control principles

Annalisa T. Taylor et al.

Summary: Active learning in robotics requires analysis and action in both abstract and physical settings. It focuses on efficient and flexible learning through continuous online deployment, posing control-oriented challenges. Suitable measures as objectives must be chosen, real-time control must be synthesized, and analyses must be produced to ensure performance and safety with limited knowledge of the environment or robot itself.

MECHATRONICS (2021)

Add to Collection

Proceedings Paper Automation & Control Systems

Multi-robot Informative Path Planning using a Leader-Follower Architecture

Gianni A. Di Caro et al.

Summary: The study focuses on using an autonomous multi-robot team to survey a large region and model the spatial map using a Gaussian Process. The team's mission is modeled as a joint informative path planning problem with a leader-follower architecture. Through computational experiments, the algorithm has shown significantly higher accuracy and good scalability in map estimation.

2021 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA 2021) (2021)

Add to Collection

Proceedings Paper Automation & Control Systems

NeuralSim: Augmenting Differentiable Simulators with Neural Networks

Eric Heiden et al.

Summary: Differentiable simulators offer a way to close the gap between simulation and reality by using efficient, gradient-based optimization algorithms. The augmentation of the novel differentiable physics engine with neural networks allows for learning nonlinear relationships and modeling effects not considered by traditional simulators. This hybrid simulation approach requires less data for training and shows improved generalization compared to entirely data-driven models.

2021 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA 2021) (2021)

Add to Collection

Proceedings Paper Automation & Control Systems

Online Informative Path Planning for Active Information Gathering of a 3D Surface

Hai Zhu et al.

Summary: This paper introduces an online informative path planning approach for active information gathering on three-dimensional surfaces using manifold Gaussian processes with geodesic kernel functions. The method plans information-gathering paths based on recent observations, respecting dynamic vehicle constraints and a total flight time budget. Results demonstrate that this approach outperforms traditional methods in reconstruction error and information-theoretic metrics, showing significant improvement in information gathering efficiency by taking spatial correlations into account.

2021 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA 2021) (2021)

Add to Collection

Proceedings Paper Automation & Control Systems

Differentiable Physics Models for Real-world Offline Model-based Reinforcement Learning

Michael Lutter et al.

Summary: A limitation of model-based reinforcement learning is the exploitation of errors in learned models. Physics-based models are better at extrapolating, while black-box models can fit complex dynamics with high fidelity. In experiments, physics-based models outperformed high-capacity function approximators in offline MBRL settings for certain tasks.

2021 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA 2021) (2021)

Add to Collection

Proceedings Paper Automation & Control Systems

Efficient Computation of Map-scale Continuous Mutual Information on Chip in Real Time

Keshav Gupta et al.

Summary: This study introduces a new hardware accelerator architecture for Mutual Information (MI) computation in exploration tasks, which significantly improves the efficiency and speed of robotic exploration. By optimizing the core and memory subsystem, real-time computation of MI for the entire map is achieved, with energy consumption at only a fraction of traditional compute platforms.

2021 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS) (2021)

Add to Collection

Proceedings Paper Computer Science, Artificial Intelligence

Active Learning in Gaussian Process State Space Model

Hon Sum Alec Yu et al.

Summary: This study investigates active learning in Gaussian Process state-space models (GPSSM) by employing mutual information as the active learning criterion. Two approaches for approximating mutual information for GPSSM given latent states are proposed and evaluated in various physical systems to actively learn the underlying nonlinear dynamics represented by the state-space model.

MACHINE LEARNING AND KNOWLEDGE DISCOVERY IN DATABASES, ECML PKDD 2021: RESEARCH TRACK, PT III (2021)

Add to Collection

Article Computer Science, Artificial Intelligence

An informative path planning framework for UAV-based terrain monitoring

Marija Popovic et al.

AUTONOMOUS ROBOTS (2020)

Add to Collection

Article Robotics

FSMI: Fast computation of Shannon mutual information for information-theoretic mapping

Zhengdong Zhang et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2020)

Add to Collection

Article Robotics

Multi-Robot Active Sensing and Environmental Model Learning With Distributed Gaussian Process

Dohyun Jang et al.

IEEE ROBOTICS AND AUTOMATION LETTERS (2020)

Add to Collection

Article Robotics

Multi-Sensor Next-Best-View Planning as Matroid-Constrained Submodular Maximization

Mikko Lauri et al.

IEEE ROBOTICS AND AUTOMATION LETTERS (2020)

Add to Collection

Article Robotics

Active sensing for motion planning in uncertain environments via mutual information policies

Ryan A Macdonald et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2019)

Add to Collection

Article Robotics

Dec-MCTS: Decentralized planning for multi-robot active perception

Graeme Best et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2019)

Add to Collection

Article Computer Science, Artificial Intelligence

Multi-modal active perception for information gathering in science missions

Akash Arora et al.

AUTONOMOUS ROBOTS (2019)

Add to Collection

Article Robotics

Variational Hilbert regression for terrain modeling and trajectory optimization

Vitor Guizilini et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2019)

Add to Collection

Article Robotics

Sampling-based incremental information gathering with applications to robotic exploration and environmental monitoring

Maani Ghaffari Jadidi et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2019)

Add to Collection

Article Computer Science, Artificial Intelligence

Localization uncertainty-aware autonomous exploration and mapping with aerial robots using receding horizon path-planning

Christos Papachristos et al.

AUTONOMOUS ROBOTS (2019)

Add to Collection

Article Robotics

Active Learning of Dynamics for Data-Driven Control Using Koopman Operators

Ian Abraham et al.

IEEE TRANSACTIONS ON ROBOTICS (2019)

Add to Collection

Article Robotics

Autonomous Exploration of Complex Underwater Environments Using a Probabilistic Next-Best-View Planner

Narcis Palomeras et al.

IEEE ROBOTICS AND AUTOMATION LETTERS (2019)

Add to Collection

Proceedings Paper Computer Science, Artificial Intelligence

Cascaded Gaussian Processes for Data-efficient Robot Dynamics Learning

Sahand Rezaei-Shoshtari et al.

2019 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS) (2019)

Add to Collection

Proceedings Paper Computer Science, Artificial Intelligence

Maximum Information Bounds for Planning Active Sensing Trajectories

Brent Schlotfeldt et al.

2019 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS) (2019)

Add to Collection

Article Robotics

Information-Guided Robotic Maximum Seek-and-Sample in Partially Observable Continuous Environments

Genevieve Flaspohler et al.

IEEE ROBOTICS AND AUTOMATION LETTERS (2019)

Add to Collection

Proceedings Paper Automation & Control Systems

Semiparametrical Gaussian Processes Learning of Forward Dynamical Models for Navigating in a Circular Maze

Diego Romeres et al.

2019 INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA) (2019)

Add to Collection

Proceedings Paper Automation & Control Systems

ChainQueen: A Real-Time Differentiable Physical Simulator for Soft Robotics

Yuanming Hu et al.

2019 INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA) (2019)

Add to Collection

Article Robotics

Variable Resolution Occupancy Mapping Using Gaussian Mixture Models

Cormac O'Meadhra et al.

IEEE ROBOTICS AND AUTOMATION LETTERS (2019)

Add to Collection

Article Computer Science, Artificial Intelligence

Receding horizon path planning for 3D exploration and surface inspection

Andreas Bircher et al.

AUTONOMOUS ROBOTS (2018)

Add to Collection

Article Robotics

Data-driven learning and planning for environmental sampling

Kai-Chieh Ma et al.

JOURNAL OF FIELD ROBOTICS (2018)

Add to Collection

Article Robotics

On the Importance of Uncertainty Representation in Active SLAM

Maria L. Rodriguez-Arevalo et al.

IEEE TRANSACTIONS ON ROBOTICS (2018)

Add to Collection

Article Robotics

Data-driven planning via imitation learning

Sanjiban Choudhury et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2018)

Add to Collection

Article Robotics

Anytime Planning for Decentralized Multirobot Active Information Gathering

Brent Schlotfeldt et al.

IEEE ROBOTICS AND AUTOMATION LETTERS (2018)

Add to Collection

Article Computer Science, Artificial Intelligence

Gaussian processes autonomous mapping and exploration for range-sensing mobile robots

Maani Ghaffari Jadidi et al.

AUTONOMOUS ROBOTS (2018)

Add to Collection

Article Computer Science, Artificial Intelligence

Revisiting active perception

Ruzena Bajesy et al.

AUTONOMOUS ROBOTS (2018)

Add to Collection

Article Robotics

Adaptive continuous-space informative path planning for online environmental monitoring

Gregory Hitz et al.

JOURNAL OF FIELD ROBOTICS (2017)

Add to Collection

Article Robotics

Adaptive informative path planning in metric spaces

Zhan Wei Lim et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2016)

Add to Collection

Article Robotics

Hilbert maps: Scalable continuous occupancy mapping with stochastic gradient descent

Fabio Ramos et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2016)

Add to Collection

Article Multidisciplinary Sciences

Robots that can adapt like animals

Antoine Cully et al.

NATURE (2015)

Add to Collection

Article Robotics

Nonmyopic View Planning for Active Object Classification and Pose Estimation

Nikolay Atanasov et al.

IEEE TRANSACTIONS ON ROBOTICS (2014)

Add to Collection

Article Robotics

Autonomous adaptive exploration using realtime online spatiotemporal topic modeling

Yogesh Girdhar et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2014)

Add to Collection

Article Robotics

Sampling-based robotic information gathering algorithms

Geoffrey A. Hollinger et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2014)

Add to Collection

Article Robotics

Active planning for underwater inspection and the benefit of adaptivity

Geoffrey A. Hollinger et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2013)

Add to Collection

Article Robotics

Optimizing waypoints for monitoring spatiotemporal phenomena

Jonathan Binney et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2013)

Add to Collection

Article Robotics

Integrated task and motion planning in belief space

Leslie Pack Kaelbling et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2013)

Add to Collection

Article Automation & Control Systems

Robots for Environmental Monitoring Significant Advancements and Applications

Matthew Dunbabin et al.

IEEE ROBOTICS & AUTOMATION MAGAZINE (2012)

Add to Collection

Article Robotics

Gaussian process occupancy maps

Simon T. O'Callaghan et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2012)

Add to Collection

Article Computer Science, Artificial Intelligence

Learning gas distribution models using sparse Gaussian process mixtures

Cyrill Stachniss et al.

AUTONOMOUS ROBOTS (2009)

Add to Collection

Article Robotics

The Belief Roadmap: Efficient Planning in Belief Space by Factoring the Covariance

Samuel Prentice et al.

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH (2009)

Add to Collection

© Peeref 2019-2024. All rights reserved.