MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202641092911 A) filed by Sr University on July 31, 2026, for An Adaptive Multi-Domain Framework For Real-Time Synthetic Speech And Audio Authentication.
Inventors include Talakoti Sathish; and Dr. Kabita Thaoroijam.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: ABSTRACT Disclosed herein is an adaptive multi-domain framework (100) for real-time synthetic speech and audio authentication that comprises an Input Interface (102), a Communication Interface (104), a Memory Unit (106), a Data Storage Unit (108), a Processing Unit (110), an Output Interface (132), and a Deployment and Integration Interface (134). The Processing Unit (110) includes an Audio Acquisition Module (112), a Signal Preprocessing Module (114), a Multi-Domain Feature Extraction Module (116), an Adaptive Feature Fusion Module (118), an Audio Authenticity Verification Module (120), an Intelligent Classification Module (122), a Confidence Score Generation Module (124), an Adaptive Calibration Module (126), a Real-Time Inference and Response Module (128), and a Data Management and Model Optimization Module (130). The framework accurately distinguishes authentic speech from synthetic speech by generating adaptive authentication decisions with corresponding confidence scores, thereby improving robustness, real-time performance, and integration across diverse voice-enabled applications.
Disclaimer: Curated by HT Syndication.
