East Point Engineering Students Develop AI System to Detect Deepfake Videos
Bengaluru, September 12, 2026: In a bid to address the growing challenge of manipulated videos and synthetic media, students of the Department of Computer Science and Engineering at East Point College of Engineering & Technology (EPCET), Bengaluru, have developed a web-based artificial intelligence system to detect deepfake videos.
The AI-powered prototype is designed to analyse uploaded videos and identify whether the content is authentic or digitally manipulated. The project was developed by Adarsh R, Chandu Kumar G, K N Siddarth and Likith Roshan H B, under the guidance of Asst. Prof. Nithyananda C R.
The system employs a hybrid AI model that combines a ResNeXt50 Convolutional Neural Network (CNN) with a Long Short-Term Memory (LSTM) model. The technology analyses facial features and patterns across multiple video frames to identify indications of manipulation.
Designed as an interactive web application, the prototype allows users to upload a video and obtain a detection result. It also includes an algorithm visualiser intended to demonstrate how the system processes video frames and extracts relevant features during detection.
According to the project team, the prototype has demonstrated the ability to identify deepfake content using the CNN-LSTM approach while providing a visual representation of the detection process. The initiative also underlines the potential of AI-based tools in promoting digital awareness and helping users critically assess the authenticity of online video content.
Commenting on the student innovation, Rajiv Gowda, CEO, East Point Group of Institutions, said deepfakes have evolved beyond being merely a technological concern and have become a matter of public trust.
“What makes this project relevant is that our students are applying AI to a problem that ordinary users, businesses and institutions increasingly face-knowing whether what they see is real. Student innovation becomes meaningful when it addresses challenges that are already shaping society,” he said.
The students have identified several areas for further development, including faster video processing, support for more video formats, improved detection accuracy and the ability to analyse videos featuring multiple faces.
The current system is being presented as a functional academic prototype and is not intended to serve as a commercial or forensic-grade video verification system. Further development and testing would be required before it could be considered for wider real-world deployment.
City Today News 9341997936
City Today News 