Learning to Live Life on the Edge: Online Learning for Data-Efficient Tactile Contour Following
This work addresses data-efficiency for robotic tactile systems, offering incremental improvements in task performance.
The paper tackles the problem of data inefficiency in tactile sensing for robotics by proposing an online learning method using a Gaussian Process Latent Variable Model, enabling successful contour following with little data and robustness to novel stimuli.
Tactile sensing has been used for a variety of robotic exploration and manipulation tasks but a common constraint is a requirement for a large amount of training data. This paper addresses the issue of data-efficiency by proposing a novel method for online learning based on a Gaussian Process Latent Variable Model (GP-LVM), whereby the robot learns from tactile data whilst performing a contour following task thus enabling generalisation to a wide variety of tactile stimuli. The results show that contour following is successful with comparatively little data and is robust to novel stimuli. This work highlights that even with a simple learning architecture there are significant advantages to be gained in efficient and robust task performance by using latent variable models and online learning for tactile sensing tasks. This paves the way for a new generation of robust, fast, and data-efficient tactile systems.