Recent advancements in 3D human pose estimation have made markerless skeletal motion recovery increasingly accurate and scalable. However, most pose estimators focus on geometric keypoint accuracy, while real-world applications in rehabilitation, sports science, ergonomics, and clinical movement analysis require biomechanical quantities that describe how the body moves, loads, and activates.
To address this, we propose BioModule, a lightweight plug-in temporal transformer that attaches downstream of any 3D pose estimator to predict biomechanical attributes from standard 17-joint 3D skeletons. BioModule is estimator-agnostic and requires no modification of the upstream pose model, allowing existing pose estimators to extend toward physically interpretable motion analysis.
To train and evaluate BioModule, we construct a large-scale aligned dataset pairing Human3.6M video and 3D keypoints with the biomechanical label space of Human3.6Mplus. We establish and verify anatomical correspondence between the coordinate systems of the two datasets, enabling frame-accurate cross-modal supervision.
Using this aligned supervision, BioModule predicts biomechanical quantities. We further benchmark BioModule across seven state-of-the-art 3D pose estimators, providing the first systematic analysis of how upstream pose estimation quality propagates to downstream biomechanical prediction fidelity. The results position BioModule as a compact, modular bridge between vision-based pose estimation and biomechanically meaningful human motion analysis.
Blogger's Review: The introduction of BioModule effectively bridges the gap between 3D pose estimation and biomechanical analysis. Its seamless integration with existing models highlights its potential applications in sports science, while the innovative aligned dataset design provides a solid foundation for future research.