Intellectual Property India Publishes Patent Application for 'System And Method For Adaptive Two Stage Retrieval Augmented Generation In Automated Incident Response' Filed by The Principal, SNS College Of Technology

MUMBAI, India, May 29 -- Intellectual Property India has published a patent application (202641043566 A) filed by The Principal, SNS College Of Technology, Coimbatore, Tamil Nadu, on April 6, for 'system and method for adaptive two stage retrieval augmented generation in automated incident response.'

Inventor(s) include Jagadeesh. B; Abishek. K. U; Shanmugavel. T; Shreyas. S; Vijayragavan. S; and Dr. Tharankumar. T.

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: "The present invention relates to an adaptive multi-modal pet behaviour analysis system comprising a user access interface (1 00), a user authentication system (11 0), a pet profile manager (120), a video upload interface (200), a video processing engine (210), an artificial intelligence analysis layer (300), a behaviour interpretation engine (330), an AI coaching layer (400), and an intelligent dashboard interface (500). Video data capturing pet behaviour is uploaded through the user interface and processed through a video frame extraction pipeline. An object detection and tracking module (31 0) identifies and tracks the spatial position of the pet within sequential video frames. A motion analysis engine (320) calculates frame-to-frame displacement to estimate activity levels and movement patterns, while a behaviour classification module (322) determines behavioural states such as resting, walking, or running. In parallel, an image captioning module (324) generates descriptive representations of observed pet actions and environmental context. Outputs from the behavioural analysis modules are synthesized by a behaviour interpretation engine (330) to estimate behavioural tendencies and mood states. An AI coaching layer (400) comprising a large language model training advisor (41 0) and a contextual advice engine (420) generates personalized training recommendations based on behavioural patterns and pet profile information. The generated insights and recommendations are presented through an intelligent dashboard interface (500) and stored within a system history database (600) to support behavioural monitoring and continuous system improvement."

Disclaimer: Curated by HT Syndication.