Policy Learning for Active Target Tracking over Continuous SE(3) Trajectories
Not on the map. Why?
Abstract (arXiv)
This paper proposes a novel model-based policy gradient algorithm for tracking dynamic targets using a mobile robot, equipped with an onboard sensor with limited field of view. The task is to obtain a continuous control policy for the mobile robot to collect sensor measurements that reduce uncertainty in the target states, measured by the target distribution entropy. We design a neural network control policy with the robot SE(3) pose and the mean vector and information matrix of the joint target distribution as inputs and attention layers to handle variable numbers of targets. We also derive the gradient of the target entropy with respect to the network parameters explicitly, allowing efficient model-based policy gradient optimization.
Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 21 of its 35 references that are in the map. First posted on arXiv 3 Dec 2022.
Builds on
21 of 35 references.
References · 21
- 2017Proximal Policy Optimization Algorithmscited by 14,441
- 2015Continuous control with deep reinforcement learningcited by 3,279
- 2015End-to-End Training of Deep Visuomotor Policiescited by 847
- 2019When to Trust Your Model: Model-Based Policy Optimizationcited by 753
- 2016Past, Present, and Future of Simultaneous Localization And Mapping: Towards the Robust-Perception Agecited by 405
- 2020Learning to Explore using Active Neural SLAMcited by 308
- 2022A Survey on Active Simultaneous Localization and Mapping: State of the Art and New Frontierscited by 62
- 2021The Sensory Neuron as a Transformer: Permutation-Invariant Neural Networks for Reinforcement Learningcited by 50
- 2020Evolutionary Population Curriculum for Scaling Multi-Agent Reinforcement Learningcited by 47
- 2019Reinforcement Learning on Variable Impedance Controller for High-Precision Robotic Assemblycited by 33
- 2020Autonomous Exploration Under Uncertainty via Deep Reinforcement Learning on Graphscited by 23
- 2018Distributed Wildfire Surveillance with Autonomous Aircraft using Deep Reinforcement Learningcited by 16
9 more
- 2022Where to Look Next: Learning Viewpoint Recommendations for Informative Trajectory Planningcited by 12
- 2020Scalable Reinforcement Learning Policies for Multi-Agent Controlcited by 10
- 2013Information Acquisition with Sensing Robots: Algorithms and Error Boundscited by 5
- 2022Learning Continuous Control Policies for Information-Theoretic Active Perceptioncited by 4
- 2019Learning Q-network for Active Information Acquisitioncited by 3
- 2021Non-Monotone Energy-Aware Information Gathering for Heterogeneous Robot Teamscited by 2
- 2021Active Exploration and Mapping via Iterative Covariance Regulation over Continuous SE(3) Trajectoriescited by 2
- 2021Active SLAM over Continuous Trajectory and Control: A Covariance-Feedback Approachcited by 1
- 2022Active Mapping via Gradient Ascent Optimization of Shannon Mutual Information over Continuous SE(3) Trajectoriescited by 1
Built on by
Citing papers per year: 2023: 1; 2024: 0; 2025: 1; 2026: 1.
3 papers.