Dynamic neuronal networks efficiently achieve classification in robotic interactions with real-world objects

TL;DR

This study demonstrates that a small, biologically inspired recurrent neuronal network can efficiently classify sensory data during robotic interactions with real-world objects, outperforming traditional statistical methods.

Contribution

The paper introduces a simple, untrained, biologically realistic neuronal network model that effectively classifies sensory data in robotic object interactions, highlighting the power of network dynamics.

Findings

The neuronal network outperformed principal component analysis in classification accuracy.

A small, untrained network achieved high performance with biologically realistic dynamics.

The approach is effective despite the network's simplicity and lack of training.

Abstract

Biological cortical networks are potentially fully recurrent networks without any distinct output layer, where recognition may instead rely on the distribution of activity across its neurons. Because such biological networks can have rich dynamics, they are well-designed to cope with dynamical interactions of the types that occur in nature, while traditional machine learning networks may struggle to make sense of such data. Here we connected a simple model neuronal network (based on the 'linear summation neuron model' featuring biologically realistic dynamics (LSM), consisting of 10 of excitatory and 10 inhibitory neurons, randomly connected) to a robot finger with multiple types of force sensors when interacting with materials of different levels of compliance. Scope: to explore the performance of the network on classification accuracy. Therefore, we compared the performance of the network output with principal component analysis of statistical features of the sensory data as well as its mechanical properties. Remarkably, even though the LSM was a very small and untrained network, and merely designed to provide rich internal network dynamics while the neuron model itself was highly simplified, we found that the LSM outperformed these other statistical approaches in terms of accuracy.