Task-driven active sensing framework applied to leaf probing Sergi Foix a,∗ , Guillem Aleny` a a , Carme Torras a a Institut de Rob`otica i Inform` atica Industrial, CSIC-UPC Llorens i Artigas 4-6, 08028 Barcelona, Spain Tel.: +34-93-4015751, Fax: +34-93-4015750 Abstract This article presents a new method for actively exploring a 3D workspace with the aim of localizing relevant regions for a given task. Our method encodes the exploration route in a multi-layer occupancy grid map. This map, together with a multiple-view estimator and a maximum-information-gain gathering ap- proach, incrementally provide a better understanding of the scene until reaching the task termination criterion. This approach is designed to be applicable to any task entailing 3D object exploration where some previous knowledge of its approximate shape is available. Its suitability is demonstrated here for a leaf probing task using an eye-in-hand arm configuration in the context of a pheno- typing application (leaf probing). Keywords: Active Perception, Next Best View, Information Gain, Search Space Reduction 1. Introduction The goal of task-driven exploration is to iteratively change the point of view so as to maximise the acquisition of information for solving a given task. We propose an algorithm that uses an information-gain criterion to compute the expected benefit of a set of candidate views, and combines it with other aspects, 5 * Corresponding author Email addresses: sfoix@iri.upc.edu (Sergi Foix), galenya@iri.upc.edu (Guillem Aleny` a), torras@iri.upc.edu (Carme Torras) Preprint submitted to Journal of L A T E X Templates February 28, 2018