ML-based tactile sensor calibration: A universal approach

TL;DR

This paper compares two tactile sensors, iCub and BioTac, analyzing their responses to external stimuli to evaluate their capabilities in force estimation and curvature recognition, using a standardized data collection and analysis approach.

Contribution

It introduces a universal calibration approach for tactile sensors by systematically studying and comparing two different sensors under identical conditions.

Findings

Both sensors perform well in different settings.

t-SNE embeddings reveal distinct stimulus representations.

Data sets enable cross-sensor calibration analysis.

Abstract

We study the responses of two tactile sensors, the fingertip sensor from the iCub and the BioTac under different external stimuli. The question of interest is to which degree both sensors i) allow the estimation of force exerted on the sensor and ii) enable the recognition of differing degrees of curvature. Making use of a force controlled linear motor affecting the tactile sensors we acquire several high-quality data sets allowing the study of both sensors under exactly the same conditions. We also examined the structure of the representation of tactile stimuli in the recorded tactile sensor data using t-SNE embeddings. The experiments show that both the iCub and the BioTac excel in different settings.