The face blurring of images plays a key role in protecting privacy. However, in computer vision, especially for the human pose estimation task, machine-learning models are currently trained, validated, and tested on original datasets without face blurring. Additionally, the accuracy of human pose estimation is of great importance for kinematic analysis. This analysis is relevant in areas such as occupational safety and clinical gait analysis where privacy is crucial. Therefore, in this study, we explore the impact of face blurring on human pose estimation and the subsequent kinematic analysis. Firstly, we blurred the subjects’ heads in the image dataset. Then we trained our neural networks using the face-blurred and the original unblurred dataset. Subsequently, the performances of the different models, in terms of landmark localization and joint angles, were estimated on blurred and unblurred testing data. Finally, we examined the statistical significance of the effect of face blurring on the kinematic analysis along with the strength of the effect. Our results reveal that the strength of the effect of face blurring was low and within acceptable limits (<1°). We have thus shown that for human pose estimation, face blurring guarantees subject privacy while not degrading the prediction performance of a deep learning model.

Accurate estimation of joint load during a lifting/lowering task could provide a better understanding of the pathogenesis and development of musculoskeletal disorders. In particular, the values of the net force and moment at the L5-S1 joint could be an important criterion to identify the unsafe lifting/lowering tasks. In this study, the joint load at L5-S1 was estimated from the motion kinematics acquired using a multi-view markerless motion capture system without force plate. The 3D human pose estimation was first obtained on each frame using deep learning. The kinematic analysis was then performed to calculate the velocity and acceleration information of each segment. Then, the net force and moment at the L5-S1 joint were calculated using inverse dynamics with a top-down approach. This estimate was compared to a reference with a bottom-up approach. It was computed using a marker-based motion capture system combined with force plates and using personalized body segment inertial parameters derived from a 3D model of the human body shape constructed for each subject using biplanar radiographs. The average differences of the estimates for force and moment among all subjects were 14.0 ± 6.9 N and 9.0 ± 2.3 Nm, respectively. Meanwhile, the mean peak value differences of the estimates were 10.8 ± 8.9 N and 11.9 ± 9.5 Nm, respectively. This study then proposed the most rigorous comparison of mechanical loading on the lumbar spine using computer vision. Further work is needed to perform such an estimation under realistic industrial conditions.