Abstract
The accurate terrain classification in real time is of great importance to an autonomous robot working in field, because the robot could avoid non-geometric hazards, adjust control scheme, or improve localization accuracy, with the aid of terrain classification. In this paper, we investigate the vibration-based terrain classification (VTC) in a dynamic environment, and propose a novel learning framework, named DyVTC, which tackles online-collected unlabeled data with concept drift. In the DyVTC framework, the exterior disagreement (ex-disagreement) and interior disagreement (in-disagreement) are proposed novely based on the feature diversity and intrinsic temporal correlation, respectively. Such a disagreement mechanism is utilized to design a pseudo-labeling algorithm, which shows its compelling advantages in extracting key samples and labeling; and consequently, the classification accuracy could be retrieved by incremental learning in a changing environment. Since two sets of features are extracted from frequency and time domain to generate disagreements, we also name the proposed method feature-temporal disagreement adaptation (FTDA). The real-world experiment shows that the proposed DyVTC could reach an accuracy of 89.5%, but the traditional time- and frequency-domain terrain classification methods could only reach 48.8% and 71.5%, respectively, in a dynamic environment.
Highlights
Robotic terrain classification refers to the process of a mobile robot classifying the terrain, on which it is traversing or will traverse, as one of the predefined classes [1]
We propose a vibration-based terrain classification framework for autonomous robots working in a dynamic environment, mainly to suppress the affect rendered by data drift, during the period that manual labels do not arrive
When the mobile robot is operating outdoors, online-collected vibration samples are fed into the pre-trained classifiers; and the classifier-output terrain predictions are fed into Bayesian filter to yield a better terrain prediction
Summary
Robotic terrain classification refers to the process of a mobile robot classifying the terrain, on which it is traversing or will traverse, as one of the predefined classes [1]. We cannot guarantee a sufficient sampling of training dataset, so it is nature to resort to the semi-supervised or unsupervised machine learning tools for VTC This idea was first proposed for safely operation of planetary exploration rovers [35]. More work that concerns the semi-supervised or unsupervised learning applying to the field of terrain classification can be found in [36,37,38] These methods could be used in static environments, where the offline training dataset and the online testing dataset are independent and identically distributed (iid). We propose a vibration-based terrain classification framework for autonomous robots working in a dynamic environment (named DyVTC), mainly to suppress the affect rendered by data drift, during the period that manual labels do not arrive.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
