MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202611089931 A) filed by Dr. Vaibhav Saran; Mr. Vaibhav Mishra; and Dr. Rajeev Kumar on July 23, 2026, for A Reliability-Adaptive Multi-Domain Forensic System And Method For Detection And Authentication Of Deepfake Digital Media.
Inventors include Dr. Vaibhav Saran; Mr. Vaibhav Mishra; Dr. Rajeev Kumar; Dr. Reena C. Jhamtani; Ms. Kirti Kalyan; Ms. Archi Gupta; Ms. Kotikalapudi Bhuvana Padma Sri; Ms. Aditi Sinha; and Ms. Mrinalinee Singh.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: [090] The present invention relates to a reliability-adaptive multi-domain forensic system and method for detection and authentication of deepfake digital media. A digital media item is acquired and preserved within a protected evidence container having cryptographically linked processing records. The media is decomposed into provenance units, spatial regions, temporal segments, acoustic segments and encoding units. Provenance-domain and content-domain forensic indicators are generated from metadata, container characteristics, acquisition residuals, compression behaviour, spatial inconsistencies, temporal behaviour, frequency characteristics, acoustic characteristics and audio-visual correspondence. A reliability value is separately determined for each available forensic domain according to media condition and feature sufficiency. The indicators are adaptively fused while retaining contradictory evidence to generate localised manipulation likelihood and uncertainty values. A manipulation-path graph associates media regions or segments with suspected processing operations. A forensic output comprising localised findings, reliability information, candidate manipulation pathways and a cryptographically verifiable audit record is thereafter generated. Accompanied Drawing [FIGS. 1-7]
Disclaimer: Curated by HT Syndication.
