Traditional biomechanical assessment is gated by costly optoelectronic motion-capture hardware, while accessible markerless alternatives such as MediaPipe and OpenPose recover joint keypoints but not the internal muscle forces and contact mechanics that drive joint degradation and soft-tissue injury. This paper details a unified end-to-end monocular pipeline that maps raw single-camera smartphone video directly into real-time internal forces, joint moments, and muscle activations at near-laboratory fidelity. The architecture integrates PromptHMR for spatial-semantic mesh reconstruction, GaitDynamics for diffusion-based generative kinetic synthesis, MuscleMimic and MuJoCo Warp for GPU-parallelized musculoskeletal simulation, low-dimensional muscle synergy priors for neurophysiological action-space constraints, SVK hyperelastic models for soft-tissue compliance at wearable interfaces, and a CNN-BiGRU-Attention predictor for real-time Knee Contact Force estimation. The document also includes a complete Product Requirements Document and 12-month engineering roadmap for the BioMotion-AI Enterprise SDK, targeting sub-16.6 ms latency on consumer mobile devices.
Kanagat et al. (Wed,) studied this question.