SitPose: Real-Time Detection of Sitting Posture and Sedentary Behavior Using Ensemble Learning With Depth Sensor

TL;DR

SitPose is a real-time system that uses depth sensors and ensemble learning to accurately detect sitting postures, aiming to improve health by reducing sedentary behavior and related disorders.

Contribution

This paper introduces a novel real-time sitting posture detection system using Kinect depth sensors and ensemble machine learning, with a new dataset and high accuracy results.

Findings

Ensemble learning achieved 98.1% F1 score in posture recognition.

The system effectively tracks 3D joint coordinates for posture analysis.

A new dataset with over 33,000 data points was created for model training.

Abstract

Poor sitting posture can lead to various work-related musculoskeletal disorders (WMSDs). Office employees spend approximately 81.8% of their working time seated, and sedentary behavior can result in chronic diseases such as cervical spondylosis and cardiovascular diseases. To address these health concerns, we present SitPose, a sitting posture and sedentary detection system utilizing the latest Kinect depth camera. The system tracks 3D coordinates of bone joint points in real-time and calculates the angle values of related joints. We established a dataset containing six different sitting postures and one standing posture, totaling 33,409 data points, by recruiting 36 participants. We applied several state-of-the-art machine learning algorithms to the dataset and compared their performance in recognizing the sitting poses. Our results show that the ensemble learning model based on the soft voting mechanism achieves the highest F1 score of 98.1%. Finally, we deployed the SitPose system based on this ensemble model to encourage better sitting posture and to reduce sedentary habits.