Optimal Deep Learning for Robot Touch: Training Accurate Pose Models of 3D Surfaces and Edges
| Autor: | John W. Lloyd, Nathan F. Lepora |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Hyperparameter
0209 industrial biotechnology Nuisance variable Computer science business.industry Deep learning Bayesian optimization Robotics 02 engineering and technology Object (computer science) Computer Science Applications 020901 industrial engineering & automation Control and Systems Engineering Robot Computer vision Artificial intelligence Electrical and Electronic Engineering business Tactile sensor |
| Zdroj: | IEEE Robotics & Automation Magazine. 27:66-77 |
| ISSN: | 1558-223X 1070-9932 |
| DOI: | 10.1109/mra.2020.2979658 |
| Popis: | This article illustrates the application of deep learning to robot touch by considering a basic yet fundamental capability: estimating the relative pose of part of an object in contact with a tactile sensor. We begin by surveying deep learning applied to tactile robotics, focusing on optical tactile sensors, which help to link touch and deep learning for vision. We then show how deep learning can be used to train accurate pose models of 3D surfaces and edges that are insensitive to nuisance variables, such as motion-dependent shear. This involves including representative motions as unlabeled perturbations of the training data and using Bayesian optimization of the network and training hyperparameters to find the most accurate models. Accurate estimation of the pose from touch will enable robots to safely and precisely control their physical interactions, facilitating a wide range of object exploration and manipulation tasks. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|