Benchmarking ML Approaches to UWB-Based Range-Only Posture Recognition for Human Robot-Interaction

TL;DR

This paper benchmarks various machine learning algorithms for human posture recognition using UWB sensors, enabling real-time robot control with high accuracy and offering a sensor-based alternative to motion sensors.

Contribution

It introduces a novel UWB-based posture recognition system and compares multiple ML models for improved accuracy in human-robot interaction applications.

Findings

UWB sensors effectively classify human postures across subjects

ML models achieve high accuracy in posture prediction

Real-time robot control demonstrated successfully

Abstract

Human pose estimation involves detecting and tracking the positions of various body parts using input data from sources such as images, videos, or motion and inertial sensors. This paper presents a novel approach to human pose estimation using machine learning algorithms to predict human posture and translate them into robot motion commands using ultra-wideband (UWB) nodes, as an alternative to motion sensors. The study utilizes five UWB sensors implemented on the human body to enable the classification of still poses and more robust posture recognition. This approach ensures effective posture recognition across a variety of subjects. These range measurements serve as input features for posture prediction models, which are implemented and compared for accuracy. For this purpose, machine learning algorithms including K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and deep Multi-Layer Perceptron (MLP) neural network are employed and compared in predicting corresponding postures. We demonstrate the proposed approach for real-time control of different mobile/aerial robots with inference implemented in a ROS 2 node. Experimental results demonstrate the efficacy of the approach, showcasing successful prediction of human posture and corresponding robot movements with high accuracy.