Online Body Schema Adaptation through Cost-Sensitive Active Learning

TL;DR

This paper introduces a cost-sensitive active learning method for online estimation of a humanoid robot's body schema, reducing movement and energy use while maintaining accuracy.

Contribution

It proposes a novel movement-efficient active learning approach for online kinematic parameter estimation in humanoid robots.

Findings

Reduces robot movement by about 50% during learning.

Maintains estimation accuracy comparable to standard active learning.

Effectively avoids uninformative configurations through occlusion modeling.

Abstract

Humanoid robots have complex bodies and kinematic chains with several Degrees-of-Freedom (DoF) which are difficult to model. Learning the parameters of a kinematic model can be achieved by observing the position of the robot links during prospective motions and minimising the prediction errors. This work proposes a movement efficient approach for estimating online the body-schema of a humanoid robot arm in the form of Denavit-Hartenberg (DH) parameters. A cost-sensitive active learning approach based on the A-Optimality criterion is used to select optimal joint configurations. The chosen joint configurations simultaneously minimise the error in the estimation of the body schema and minimise the movement between samples. This reduces energy consumption, along with mechanical fatigue and wear, while not compromising the learning accuracy. The work was implemented in a simulation environment, using the 7DoF arm of the iCub robot simulator. The hand pose is measured with a single camera via markers placed in the palm and back of the robot's hand. A non-parametric occlusion model is proposed to avoid choosing joint configurations where the markers are not visible, thus preventing worthless attempts. The results show cost-sensitive active learning has similar accuracy to the standard active learning approach, while reducing in about half the executed movement.