Trajectory Optimization using Reinforcement Learning for Map Exploration

By Thomas Kollar, Nicholas A. Roy. 2008

163 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Reinforcement learning in robotics: A survey (2013, background): “It has been demonstrated in learning walking controllers and in finding near-optimal trajectories for map exploration (Kollar and Roy, 2008).”
IQA: Visual Question Answering in Interactive Environments (2017, background): “Reinforcement Learning (RL): RL algorithms have been employed in a wide range of problems including locomotion [33], obstacle detection [44] and autonomous flight [34, 59].”
Active vision in robotic systems: A survey of recent developments (2011, methodology): “Learning and expert system Interactive learning (Grewe and Kak 1995) or reinforcement learning (Kollar and Roy 2008) is frequently used for active recognition, localization, planning, and modeling (Wang et al. 2008).”
Learning to Explore using Active Neural SLAM (2020, background): “Most such works frame this problem as a Partially Observable Markov Decision Process (POMDP) that are approximately solved (Martinez-Cantin et al., 2009; Kollar and Roy, 2008), and or seek to find a sequence of actions that minimizes uncertainty of maps (Stachniss et al.”
Learning to Look Around: Intelligently Exploring Unseen Environments for Unknown Tasks (2017, background): “Active visual localization and mapping: Active visual SLAM aims to limit samples needed to densely reconstruct a 3D environment using geometric methods [13, 34, 35, 44, 60].”
A Bayesian exploration-exploitation approach for optimal online sensing and planning with a visually guided mobile robot (2009, methodology): “However, since the dynamic model is non-differentiable, one cannot use gradient-based stoc hastic approximations [Konda and Tsitsiklis (2003); Singh et a l (2005)] or policy gradient methods [Baxter and Bartlett (20 01); Peters and Schaal (2008a,b)] to update the parameters of the policy. Instead, w”
Learning to soar: Resource-constrained exploration in reinforcement learning (2015, methodology): “…maximizing properties of the information matrix such as maximizing the minimum eigenvalue in E-optimality, or minimizing various properties of the covariance matrix, such as the trace or determinant for A-and D-optimality, respectively, as applied in Kollar and Roy (2008) and Binney et al. (2013).”
Learning Exploration Policies for Navigation (2019, background): “Kollar & Roy (2008) assume access to the ground-truth map and learn an optimized trajectory that maximizes the accuracy of the SLAM-derived map.”