MUMBAI, India, Feb. 27 -- Intellectual Property India has published a patent application (202641018822 A) filed by BVRIT Hyderabad College Of Engineering For Women; G Purnachandrarao; and Perabattula Rajeshkumar, Hyderabad, Telangana, on Feb. 19, for 'an intelligent monitoring and control system using artificial intelligence for adaptive deepfake recognition.'
Inventor(s) include Bvrit College Engineering For Women; G Purnachandrarao; and Perabattula Rajeshkumar.
The application for the patent was published on Feb. 27, under issue no. 09/2026.
According to the abstract released by the Intellectual Property India: "The present invention discloses an intelligent monitoring and control system for adaptive deepfake recognition using artificial intelligence and deep learning techniques. The system is configured to automatically detect, analyze, and classify manipulated facial images within digital media through a structured, multi stage processing framework. Input images are acquired from diverse sources and subjected to preprocessing operations including resizing, normalization, noise reduction, and illumination correction to ensure uniform representation and stable performance. A face detection and alignment module identifies valid facial regions and performs landmark based alignment to reduce variations in pose, scale, and facial expression. The aligned facial data is processed by a deep feature extraction module based on residual and inception inspired neural network architectures, generating robust and discriminative facial embeddings. These embeddings capture both global and fine grained facial characteristics essential for distinguishing authentic images from synthetically generated or manipulated media. A face verification and similarity measurement module compares the extracted embeddings using distance metrics and predefined thresholds to determine authenticity. A decision control module produces a final classification output indicating real or fake media with associated confidence levels. The invention operates through an adaptive and modular architecture that enables scalable deployment, efficient computation, and near real time performance. The integrated pipeline improves detection accuracy, reduces false acceptance and false rejection rates, and enhances robustness against high quality synthetic media and adversarial artifacts. The system is applicable to digital forensics, biometric authentication, cybersecurity monitoring, surveillance systems, and online content moderation platforms. By providing an automated, reliable, and high accuracy solution, the invention strengthens digital media integrity, security, and trust in modern communication environments."
Disclaimer: Curated by HT Syndication.
