ITSC 2025 Paper Abstract

Paper FR-LM-T44.6

Cui, Shiyan (Technical University of Munich), Hayashi, Hiroaki (Waseda University), Kamezaki, Mitsuhiro (The University of Tokyo)

Development of a Driving Posture Assessment System for Posture Standardization Using a Single In-Cabin Camera

Scheduled for presentation during the Regular Session "S44a-Human Factors and Human Machine Interaction in Automated Driving" (FR-LM-T44), Friday, November 21, 2025, 12:10−12:30, Currumbin

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on October 18, 2025

Keywords Driver Behavior Monitoring and Feedback Systems for Semi-autonomous Vehicles, Human-Machine Interaction Systems for Enhanced Driver Assistance and Safety, Real-time Object Detection and Tracking for Dynamic Traffic Environments

Abstract

Poor driving posture can impair musculoskeletal health and reduce situational awareness by narrowing the field of view. Although previous research has defined standard driving positions through ergonomic investigations, few studies have used non-contact sensors for real-time monitoring. To address this issue, we propose a driver posture assessment system using a single in-cabin camera. Our method integrates YOLOX for driver detection and HRNet for skeletal keypoint estimation, combined with the Segment Anything Model (SAM) for body and seat segmentation. From these outputs, key posture angles are calculated from the geometry of the detected keypoints and by fitting the driver’s silhouette relative to the seat. Each approach exhibits strengths and weaknesses; therefore, a Kalman filter is applied to fuse the two types of angle measurements, resulting in more stable and accurate posture estimates. We evaluate the proposed system by comparing its output to posture labels annotated based on established ergonomic studies, and the results show that the fused approach closely aligns with human assessments. We confirm that this continuous monitoring system can assist drivers in maintaining proper posture during regular driving.