• Gauri Bharat Pawar
• Dr. R.N Atole
• Nisha Sunil Khalate
• Vaishnavi Jyotiram Suryawanshi
• Mugdha Mahendra Kashid

In the digital era, video has emerged as a leading medium for information sharing. However, the increasing ease of manipulating video content poses serious challenges to authenticity, especially in law enforcement, journalism, and national security. This paper introduces an AI-based dual-detection system to identify both deepfakes and frame-level video tampering.The system utilizes two deep learning models: a CNN for detecting deepfakes and a ResNet18 model for classifying tampered frames, including splicing, cloning, and inpainting. Video frames are extracted and analyzed in real-time, with results presented visually.Developed using Python and Flask, the platform includes user authentication via SQLite and a responsive HTML interface for seamless interaction. Results are shown with frame-wise logs and video previews. Experimental results demonstrate high accuracy and reliability. This tool offers scalable, real-time video integrity verification with future scope for GPU support and audio forgery detection.

Deepfake Detection, Video Tampering, CNN, ResNet18, PyTorch, Flask, Real-time Detection, Forensics.

"AI-Powered Video Tampering and Deepfake Detection System using deep learning ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 5, page no.f166-f173, May-2025, Available :http://www.jetir.org/papers/JETIR2505615.pdf

"AI-Powered Video Tampering and Deepfake Detection System using deep learning ", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 5, page no. ppf166-f173, May-2025, Available at : http://www.jetir.org/papers/JETIR2505615.pdf