Object Recognition, Dynamic Contact Simulation, Detection, and Control of the Flexible Musculoskeletal Hand Using a Recurrent Neural Network with Parametric Bias

TL;DR

This paper presents a recurrent neural network with parametric bias to model a flexible musculoskeletal hand, enabling object recognition, contact detection, and control, while adapting to deterioration and variability over time.

Contribution

It introduces a unified neural network approach for multiple hand functions that adapts over time, addressing modeling challenges of flexible musculoskeletal hands.

Findings

Effective recognition of grasped objects

Successful contact simulation and detection

Adaptability to deterioration and initialization variability

Abstract

The flexible musculoskeletal hand is difficult to modelize, and its model can change constantly due to deterioration over time, irreproducibility of initialization, etc. Also, for object recognition, contact detection, and contact control using the hand, it is desirable not to use a neural network trained for each task, but to use only one integrated network. Therefore, we develop a method to acquire a sensor state equation of the musculoskeletal hand using a recurrent neural network with parametric bias. By using this network, the hand can realize recognition of the grasped object, contact simulation, detection, and control, and can cope with deterioration over time, irreproducibility of initialization, etc. by updating parametric bias. We apply this study to the hand of the musculoskeletal humanoid Musashi and show its effectiveness.