Semi-supervised Gated Recurrent Neural Networks for Robotic Terrain Classification

TL;DR

This paper demonstrates that gated recurrent neural networks can effectively classify terrains for legged robots using both supervised and semi-supervised learning, outperforming frequency-domain methods and enabling adaptive behaviors.

Contribution

It introduces a semi-supervised Gated Recurrent Neural Network approach for terrain classification, improving accuracy with limited labeled data and raw unlabelled data.

Findings

Time-domain classifiers outperform frequency-domain classifiers.

Semi-supervised learning significantly improves classification accuracy.

Robust terrain classification enables adaptive robot behaviors.

Abstract

Legged robots are popular candidates for missions in challenging terrains due to the wide variety of locomotion strategies they can employ. Terrain classification is a key enabling technology for autonomous legged robots, as it allows the robot to harness their innate flexibility to adapt their behaviour to the demands of their operating environment. In this paper, we show how highly capable machine learning techniques, namely gated recurrent neural networks, allow our target legged robot to correctly classify the terrain it traverses in both supervised and semi-supervised fashions. Tests on a benchmark data set shows that our time-domain classifiers are well capable of dealing with raw and variable-length data with small amount of labels and perform to a level far exceeding the frequency-domain classifiers. The classification results on our own extended data set opens up a range of high-performance behaviours that are specific to those environments. Furthermore, we show how raw unlabelled data is used to improve significantly the classification results in a semi-supervised model.