Print Email Facebook Twitter Active Classification of Moving Targets With Learned Control Policies Title Active Classification of Moving Targets With Learned Control Policies Author Serra Gomez, A. (TU Delft Learning & Autonomous Control) Montijano, Eduardo (Universidad de Zaragoza) Böhmer, J.W. (TU Delft Algorithmics) Alonso-Mora, J. (TU Delft Learning & Autonomous Control) Date 2023 Abstract In this paper, we consider the problem where a drone has to collect semantic information to classify multiple moving targets. In particular, we address the challenge of computing control inputs that move the drone to informative viewpoints, position and orientation, when the information is extracted using a “black-box” classifier, e.g., a deep learning neural network. These algorithms typically lack of analytical relationships between the viewpoints and their associated outputs, preventing their use in information-gathering schemes. To fill this gap, we propose a novel attention-based architecture, trained via Reinforcement Learning (RL), that outputs the next viewpoint for the drone favoring the acquisition of evidence from as many unclassified targets as possible while reasoning about their movement, orientation, and occlusions. Then, we use a low-level MPC controller to move the drone to the desired viewpoint taking into account its actual dynamics. We show that our approach not only outperforms a variety of baselines but also generalizes to scenarios unseen during training. Additionally, we show that the network scales to large numbers of targets and generalizes well to different movement dynamics of the targets. Subject Machine learning for robot controlreactive and sensor-based planningsurveillance robotic systems To reference this document use: http://resolver.tudelft.nl/uuid:0b54f0b3-82ad-4b43-8ed4-5ae09f2d8e53 DOI https://doi.org/10.1109/LRA.2023.3271508 Embargo date 2023-10-28 ISSN 2377-3766 Source IEEE Robotics and Automation Letters, 8 (6), 3717-3724 Bibliographical note Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public. Part of collection Institutional Repository Document type journal article Rights © 2023 A. Serra Gomez, Eduardo Montijano, J.W. Böhmer, J. Alonso-Mora Files PDF Active_Classification_of_ ... licies.pdf 2.39 MB Close viewer /islandora/object/uuid:0b54f0b3-82ad-4b43-8ed4-5ae09f2d8e53/datastream/OBJ/view