MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202641124505 A) filed by Mlr Institute Of Technology; and Marri Laxman Reddy Institute Of Technology And Management on October 17, 2026, for Ai-Based Deepfake Detection System For Multimodal Digital Media.
Inventors include Dr. Ajmeera Kiran; Mr. G. Jignesh Kumar; Ms. E. Sowmya; and Mr. P. Ajay.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: The present invention relates to an artificial intelligence-based deepfake detection system configured to identify potentially manipulated digital images, videos, and audio recordings using deep learning techniques. The system comprises a media input module, a preprocessing module, a feature extraction module, a deepfake classification module, a confidence assessment module, and an output interface. The media input module receives digital media for analysis, and the preprocessing module normalizes the input data into a format suitable for processing. The feature extraction module derives spatial, temporal, facial, textural, and acoustic features from the input media to identify characteristics associated with synthetic or manipulated content. The deepfake classification module processes the extracted features using trained machine learning models to classify the input media according to the likelihood of manipulation. For video inputs, the system analyzes frame-level features and temporal inconsistencies, while for audio inputs, it evaluates acoustic patterns and signal characteristics. The system generates a classification result, an associated confidence score, and, where supported by the analysis, indicators of potentially manipulated regions or segments. The results are presented through a user interface to support media verification and further examination. The system may further incorporate model updating mechanisms to accommodate newly identified manipulation techniques. The invention facilitates automated multimedia analysis for digital content verification, cybersecurity, and detection of potentially fraudulent or misleading media.
Disclaimer: Curated by HT Syndication.
