MUMBAI, India, May 29 -- Intellectual Property India has published a patent application (202641062160 A) filed by Koneru Lakshmaiah Education Foundation, Guntur, Andhra Pradesh, on May 15, for 'hybrid deep learning system for copy-move forgery detection in images.'
Inventor(s) include Ravula Jyothsna; and Nilu Singh.
The application for the patent was published on May 29, under issue no. 22/2026.
According to the abstract released by the Intellectual Property India: "The present invention discloses a hybrid deep learning-based system (100) for detecting and localizing copy-move forgery in digital images. The system comprises an image input module (102), a preprocessing module (104), a Generative Adversarial Network module (106), a Vision Transformer module (108), a graph-based learning module (110), a Capsule Network module (112), a feature fusion and classification module (114), and an output generation module (116). The disclosed system integrates GAN-based forged image generation, transformer-based contextual feature extraction, graph-based relational learning, and capsule-based hierarchical feature preservation for improved forgery detection accuracy and localization capability. The system is robust against scaling, rotation, compression, blurring, illumination variation, and noise attacks, thereby enabling reliable digital image authentication for forensic, cybersecurity, surveillance, and legal evidence verification applications."
Disclaimer: Curated by HT Syndication.
