MUMBAI, India, May 29 -- Intellectual Property India has published a patent application (202641061625 A) filed by Vellore Institute Of Technology, Vellore, Tamil Nadu, on May 14, for 'system and method for physics-informed spatial-temporal graph-based human pose refinement and biomechanical analysis.'
Inventor(s) include Dr. Jerart Julus L; and Adheesh Dubey.
The application for the patent was published on May 29, under issue no. 22/2026.
According to the abstract released by the Intellectual Property India: "A system and method for physics-informed refinement of human pose data using a spatial-temporal graph-based framework are disclosed. The system receives pose data representing body joint positions and constructs a skeletal graph representation with joints as nodes and connections as edges. A spatial-temporal graph-based neural network processes the graph to model spatial relationships and temporal dependencies across pose sequences. A constraint integration module incorporates biomechanical constraints, including bone length consistency and joint range-of-motion limits, using a composite loss function to ensure anatomical validity. The system refines pose sequences to reduce temporal inconsistencies and generate stable and physically plausible outputs. The method supports real-time execution in client-side computing environments, enabling efficient and low-latency pose refinement. The invention improves accuracy, temporal stability, and biomechanical consistency of pose estimation systems."
Disclaimer: Curated by HT Syndication.
