MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202641090505 A) filed by Manipal Academy Of Higher Education on July 24, 2026, for A System And Method For Privacy-Preserving Edge-Based Threat Detection In A Domestic Environment.
Inventors include Gonugondla Veera Manikanta; and Swathi Tangi.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: The present invention provides a system and method for privacy-preserving edge-based threat detection. A host processor 103 receives video frames from a camera sensor 101 over a serial camera interface. A neural processing unit 105 hosts a primary detection neural network 107 and a secondary verification neural network 109. The primary detection neural network 107 processes each frame to generate a bounding box, class label, and confidence value for a detected region of interest. A cropped image region is forwarded to a bounded queue and evaluated by the secondary verification neural network 109 to produce a binary safe-versus-threat classification and a depth-variance measure. The host processor 103 discards cropped image regions when the queue reaches capacity, assigns processing cores by core affinity, generates an alert signal upon confirmed classification over consecutive frames, and restricts transmission of video frames and inference outputs to any server outside a user-owned trust boundary. (FIG.2)
Disclaimer: Curated by HT Syndication.
