Probabilistic Discriminative Models Address the Tactile Perceptual Aliasing Problem

TL;DR

This paper addresses tactile perceptual aliasing in robotic touch by proposing a probabilistic discriminative model that improves identification of ambiguous stimuli compared to traditional neural networks.

Contribution

It introduces a mixture density network approach to detect and handle perceptual aliasing in tactile data, enhancing prediction accuracy under ambiguity.

Findings

Mixture density network outperforms deep neural networks on aliased data.

Discriminative models struggle with ambiguous tactile stimuli.

Uncertain predictions reveal sources of perceptual ambiguity.

Abstract

In this paper, our aim is to highlight Tactile Perceptual Aliasing as a problem when using deep neural networks and other discriminative models. Perceptual aliasing will arise wherever a physical variable extracted from tactile data is subject to ambiguity between stimuli that are physically distinct. Here we address this problem using a probabilistic discriminative model implemented as a 5-component mixture density network comprised of a deep neural network that predicts the parameters of a Gaussian mixture model. We show that discriminative regression models such as deep neural networks and Gaussian process regression perform poorly on aliased data, only making accurate predictions when the sources of aliasing are removed. In contrast, the mixture density network identifies aliased data with improved prediction accuracy. The uncertain predictions of the model form patterns that are consistent with the various sources of perceptual ambiguity. In our view, perceptual aliasing will become an unavoidable issue for robot touch as the field progresses to training robots that act in uncertain and unstructured environments, such as with deep reinforcement learning.