MUMBAI, India, July 13 -- Intellectual Property India has published a patent application (202641081568 A) filed by Vijay Krishna Maddukuri; K. S. N. Prasad; S. V. V. D. Venu Gopal; K. Viswaprasad; A. Revathi; M. Vinay Kiran; and Sasi Institute Of Technology & Engineering on July 02, 2026, for Deep Learning-Based Voice Authentication System For Secure Digital Access.
Inventors include Vijay Krishna Maddukuri; K. S. N. Prasad; S. V. V. D. Venu Gopal; K. Viswaprasad; A. Revathi; and M. Vinay Kiran.
The application for the patent was published on July 10, 2026, under issue no. 28/2026.
Abstract: The present invention relates to a Deep Learning-Based Voice Authentication System for Secure Digital Access that provides a secure and efficient biometric authentication mechanism using voice characteristics. The system captures a user's voice sample, performs speech preprocessing operations, and utilizes an ECAPA-TDNN deep learning model to extract distinctive speaker embeddings. The extracted embeddings are stored during enrollment and subsequently compared with newly generated embeddings during authentication using a cosine similarity matching technique. Based on the similarity score, the system grants or denies access to protected digital resources. The invention eliminates the dependence on conventional password-based authentication methods and enhances security against unauthorized access, password theft, and credential misuse. The proposed framework supports real-time authentication through a web-based interface and incorporates advanced speech processing and deep learning technologies to improve accuracy, reliability, and user convenience. The system is suitable for deployment in banking, healthcare, educational, enterprise, and other digital platforms requiring secure identity verification and access control.
Disclaimer: Curated by HT Syndication.
