A Database of Vibratory Signals from Free Haptic Exploration of Natural Material Textures and Perceptual Judgments (ViPer): Analysis of Spectral Statistics.

Abstract

We recorded vibratory patterns elicited by free haptic exploration of a large set of natural textures with a steel tool tip. Vision and audio signals during the exploration were excluded. After the exploration of each sample, participants provided judgments about its perceptual attributes and material category. We found that vibratory signals can be approximated by a single parameter in the temporal frequency domain, in a similar way as we can describe the spatial frequency spectrum of natural images. This parameter varies systematically between material categories and correlates with human perceptual judgements. It provides an estimate of the spectral composition of the power spectra which is highly correlated with the differential activity of the Rapidly Adapting (RA) and Pacinian Corpuscle (PC) afferents.