MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202641090746 A) filed by Bgs College Of Engineering And Technology Bgscet on July 25, 2026, for A System And Method For Identifying Manipulated Or Ai-Generated Digital Images.
Inventors include Ravikumar Guralamata Krishnegowda; Govindaiah Thimma Raju; Madhu Patil; Chandrakala Hassan Lakshmegowda; Kirankumar Nuggalapura Appoji Gowda; Santhosh Siddapur Mallesha; Tharun Kodihalli Lakshmana; and Jagadish Kumar Narayanappa.
The application for the patent was published on July 31, 2026, under issue no. 31/2026.
Abstract: Existing image authentication techniques often exhibit reduced reliability against sophisticated manipulated and AI-generated images due to dependence on isolated forensic analysis. The present disclosure provides a computer-implemented method and system (106, 202) for identifying manipulated or AI-generated digital images by receiving an input image, performing preprocessing through an image acquisition module (204), extracting spatial, frequency-domain, metadata, and statistical forensic features using a feature extraction and analysis module (206), generating a unified feature representation through a feature fusion and representation module (208), and determining image authenticity using a classification and decision module (210). The system further localizes manipulated regions, generates explainability visualizations, confidence scores, and a structured forensic report through an output interface (212). The disclosed architecture improves robustness, detection accuracy, interpretability, and adaptability across diverse image manipulation techniques.
Disclaimer: Curated by HT Syndication.
