MUMBAI, India, June 30 -- Intellectual Property India has published a patent application (202641073778 A) filed by Vellore Institute Of Technology on June 14, 2026, for Reinforcement-Threshold Controlled Spatiotemporal Gesture Interpretation System Using Adaptive Memory Engine.
Inventors include Viswanathan P; and Harapriya Kar.
The application for the patent was published on June 26, 2026, under issue no. 26/2026.
Abstract: REINFORCEMENT-THRESHOLD CONTROLLED SPATIOTEMPORAL GESTURE INTERPRETATION SYSTEM USING ADAPTIVE MEMORY ENGINE The present invention relates to a reinforcement-threshold controlled spatiotemporal intelligence system for real-time multilingual gesture interpretation. The system receives continuous gesture image sequences and extracts multimodal landmarks associated with hand movements, facial expressions, and body posture for generating unified spatiotemporal gesture representations. A reinforcement-threshold control module dynamically regulates attention distributions and adaptive memory updates according to gesture complexity during continuous gesture processing. The system further comprises a Harprigesthra Adaptive Memory Engine having temporal and spatiotemporal memory states configured to preserve short-term gesture transitions and hierarchical and long-range gesture dependencies using contextual loss feedback. An adaptive mode selection module allocates processing modes according to gesture processing conditions. The interpreted gesture sequences are converted into multilingual text or speech outputs supporting Indian Sign Language, American Sign Language, and Chinese Sign Language for low-latency assistive communication applications.
Disclaimer: Curated by HT Syndication.
