MUMBAI, India, Jan. 9 -- Intellectual Property India has published a patent application (202521130797 A) filed by Symbiosis International, Pune, Maharashtra, on Dec. 23, 2025, for 'a novel system for generating augmented 3d human pose data from minimal input.'
Inventor(s) include Dr. Vijayshri Khedkar; Yash Parkhi; Vrushant Mukherjee; and Mr. Anthony Augustine.
The application for the patent was published on Jan. 9, under issue no. 02/2026.
According to the abstract released by the Intellectual Property India: "The present invention relates to a system (100) for generating personalized synthetic training data for human pose estimation from minimal input images (108). A Calibration Subsystem (102) uses an SMPL-X regressor (110) to extract shape parameters beta (112) creating a Digital Mannequin (114). A Geometry-Only Synthetic Data Factory (104) animates the mannequin (114) through a Geometric Domain Randomization Pipeline (116) generating synthetic frames (202). A Target-in-the-Loop Validation Module (200) processes frames (202) through a Lightweight Keypoint Extraction Model (206) producing Detected Keypoints (208) paired with True Labels (204). An Edge-Optimized Training Module (106) trains a Temporal Classification Network (210) and applies quantization (301) creating a Quantized Classifier (308). The system deploys on Embedded Single-Board Computers (300) processing Live Video Frames (302) through a Runtime Keypoint Detector (304), Temporal Sliding Window (306), and Quantized Classifier (308) providing feedback via Audio-Visual Feedback Module (310) without specialized sensors."
Disclaimer: Curated by HT Syndication.
