Impact Factor: 7.883
Submit Paper
Volume 11, Issue 10 (October 2025)

Enhanced Deepfake Detection Using Preprocessed Video Frames And Convolution Neural Networks

Impact Factor
7.883
Call For Paper
Volume 12 Issue 07

July 2026

Download Paper Format
Paper Format
Copyright Form
Copy Right Format
Under License Of
IJSART License
Creative Commons Attribution-NonCommercial
4.0 International License
Share on:

Author(s)

R.Mohana Brintha R. Pavithra A. Alagar

Abstract

Digital Security And Authenticity Are Being Threatened By Deepfake Technology, Which Is Powered By Sophisticated Generative Models. A Convolutional Neural Network (CNN) Model Trained On Facial Data Taken From Video Frames Is Used In This Study To Demonstrate A Deep Learning-based Method For Identifying Deepfake Films. Prior To Classification, The System Preprocesses Video Data By Normalizing And Shrinking Frames. Real-time Video Uploading And Prediction Are Made Possible By A Gradio-based User Interface, Which Indicates The Likelihood Of Authentic Or Fraudulent Footage. The Model's Ability To Counteract Synthetic Media And Guarantee Trust In Visual Content Across Digital Communication Channels Is Highlighted By Experimental Results That Show Dependable Detection Performance.

Keywords

Deepfake Detection Convolutional Neural Network (CNN) Video Analysis Machine Learning Deep Learning Face Recognition Image Preprocessing Artificial Intelligence Feature Extraction Classification Gradio Interface Model Prediction Frame Processi

Paper ID

IJSARTV11I10104179

Publication Date

October 24, 2025

Research Area

Computer Science And Engineering

Download Full Article

Submit Your Paper to IJSART

Join the global research community with IJSART. Submit your paper, share your work, and gain worldwide recognition!

Submit Your Paper