MUMBAI, India, July 24 -- Intellectual Property India has published a patent application (202641085161 A) filed by Koneru Lakshmaiah Education Foundation on July 11, 2026, for Streaming-Based Robust Voice Authentication Framework For Smart Home Safety Using Deep Neural Networks.
Inventors include Sasmita Kumari Pradhan; and Dr. Suryakanth V. Gangashetty.
The application for the patent was published on July 17, 2026, under issue no. 29/2026.
Abstract: The present invention relates to a streaming-based voice authentication framework for enhancing security in smart home environments using an end-to-end deep neural network. The framework acquires streaming voice signals, preprocesses the signals using a modified digital wavelet transform to reduce noise and improve signal quality, and extracts frequency, energy, and temporal variation features to generate discriminative feature vectors. The extracted features are processed by a trained deep neural network to compute a Frequency Support Measure (FSM), an Energy Support Measure (ESM), a Variational Support Measure (VSM), and a User Support Measure (USM) for authenticating users. Based on the authentication outcome, access to smart home devices and services is selectively granted or denied. The framework provides improved authentication accuracy, reduced false authentication rates, enhanced computational efficiency, and robust real-time performance for secure smart home applications.
Disclaimer: Curated by HT Syndication.
