MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202541039069 A) filed by Murugalakshmi P on April 23, 2025, for Skill Trade Platform.
Inventors include Murugalakshmi P; Shreeja N; Hariharan K; Deepak Raj T; and Tamilselvan A.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: ABSTRACT The rise of deepfake technology, driven by powerful deep learning techniques, has made it possible to create highly convincing fake videos and images of individuals. These synthetic faces are often indistinguishable to. the human eye, posing serious risks in areas like politics, cybersecurity, and social media. This project aims to address this growing concern by developing an automated deepfake face detection system. Using deep neural networks, particularly Convolutional Neural Networks (CNNs), the system is trained to identify artifacts, inconsistencies, and unnatural facial movements commonly found in deepfakes. Public datasets such as FaceForensics++, Celeb-OF, and DFDC are used for training and evaluation. The proposed model achieves strong accuracy and performance, making it a promising tool for realworld applications in digital forensics and media verification. This research highlights the critical role of AI in safeguarding against misinfonnation and ensuring the authenticity of visual content. ABSTRACT The rapid advancement of deepfake technology has raised significant concerns regarding the authenticity of digital media, especially in the context of misinformation, identity theft, and fraud. This project presents a deepfake face detection system that utilizes state-of-the-art deep learning techniques, particularly Convolutional Neural Networks (CNNs), to identify manipulated facial images and videos. The system analyzes facial features such as texture, lighting, and unnatural movement patterns, which are often present in deepfakes but absent in genuine media. By preprocessing input images and videos, extracting facial regions, and applying a trained CNN model, the system efficiently distinguishes between real and fake content. The model is trained on publicly available datasets, including FaceForensics++ and Celeb-OF, ensuring robustness across different types of deepfake techniques. Additionally, the system supports real-time detection and outputs a confidence score that indicates the likelihood of content being manipulated. The developed solution can be integrated into various platforms, such as social media, digital forensics, and content moderation tools, to help preserve the integrity of digital media. This project aims to provide an automated, scalable, and effective method for combating the growing threat of deepfake technology.
Disclaimer: Curated by HT Syndication.
