Leveraging Robotic Prior Tactile Exploratory Action Experiences For Learning New Objects's Physical Properties

TL;DR

This paper presents a method for robots to transfer prior tactile experiences from previous objects to improve learning new objects' physical properties, achieving over 10% accuracy improvement and robustness against irrelevant knowledge.

Contribution

The work introduces a transfer learning approach for tactile exploration in robots, enabling improved object property recognition using prior tactile experiences.

Findings

Over 10% improvement in discrimination accuracy with prior knowledge

25% accuracy gain with only one training sample

Robustness against irrelevant tactile knowledge transfer

Abstract

Reusing the tactile knowledge of some previously-explored objects helps us humans to easily recognize the tactual properties of new objects. In this master thesis, we enable arobotic arm equipped with multi-modal artificial skin, like humans, to actively transfer the prior tactile exploratory action experiences when it learns the detailed physical properties of new objects. These prior tactile experiences are built when the robot applies the pressing, sliding and static contact movements on objects with different action parameters and perceives the tactile feedbacks from multiple sensory modalities. Our method was systematically evaluated by several experiments. Results show that the robot could consistently improve the discrimination accuracy by over 10% when it exploited the prior tactile knowledge compared with using no transfer method, and 25% when it used only one training sample. The results also show that the proposed method was robust against transferring irrelevant prior tactile knowledge.