Organization and Understanding of a Tactile Information Dataset TacAct During Physical Human-Robot Interactions

TL;DR

This paper introduces TacAct, a comprehensive tactile dataset from human-robot interactions, and demonstrates its potential for training neural networks to classify touch actions, advancing tactile intelligence in service robots.

Contribution

The paper presents a new large-scale tactile dataset TacAct, including data collection, organization, preliminary analysis, and validation using neural networks, to enhance tactile perception in robots.

Findings

The dataset contains 24,000 touch actions from 50 subjects.

Preliminary analysis shows the data's diversity and richness.

Neural network classification achieves promising accuracy on touch types.

Abstract

Advanced service robots require superior tactile intelligence to guarantee human-contact safety and to provide essential supplements to visual and auditory information for human-robot interaction, especially when a robot is in physical contact with a human. Tactile intelligence is an essential capability of perception and recognition from tactile information, based on the learning from a large amount of tactile data and the understanding of the physical meaning behind the data. This report introduces a recently collected and organized dataset "TacAct" that encloses real-time pressure distribution when a human subject touches the arms of a nursing-care robot. The dataset consists of information from 50 subjects who performed a total of 24,000 touch actions. Furthermore, the details of the dataset are described, the data are preliminarily analyzed, and the validity of the collected information is tested through a convolutional neural network LeNet-5 classifying different types of touch actions. We believe that the TacAct dataset would be more than beneficial for the community of human interactive robots to understand the tactile profile under various circumstances.