https://github.com/giuleo129/deepfake_detection

A collection of projects exploring DeepFake detection using pretrained neural networks with fine-tuning and SVM classification on Fourier-transformed features.
https://github.com/giuleo129/deepfake_detection

deep-learning deep-neural-networks deepfake deepfake-detection feature-extraction fourier-transform gan jupyter-notebook pretrained-models python pytorch svm-classifier

Last synced: 2 days ago
JSON representation

A collection of projects exploring DeepFake detection using pretrained neural networks with fine-tuning and SVM classification on Fourier-transformed features.

• Host: GitHub
• URL: https://github.com/giuleo129/deepfake_detection
• Owner: giuleo129
• Created: 2025-02-11T13:24:59.000Z (2 days ago)
• Default Branch: main
• Last Pushed: 2025-02-11T13:59:02.000Z (2 days ago)
• Last Synced: 2025-02-11T14:35:24.499Z (2 days ago)
• Topics: deep-learning, deep-neural-networks, deepfake, deepfake-detection, feature-extraction, fourier-transform, gan, jupyter-notebook, pretrained-models, python, pytorch, svm-classifier
• Language: Jupyter Notebook
• Homepage:
• Size: 6.38 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
🔍 DeepFake Analysis & Detection

dataset link: https://iplab.dmi.unict.it/mfs/Deepfakes/PaperMDPI2022/

This repository contains projects dedicated to the study and detection of DeepFake images using various machine learning techniques. The research focuses on two main approaches:

Pretrained Neural Networks & Fine-Tuning:

1- Leveraging state-of-the-art models (e.g., AlexNet, GoogleNet, ViT) to classify DeepFake images.

2- Exploring the potential of fine-tuning to improve model performance.

Fourier-Based SVM Classification

1- Extracting frequency-domain features from DeepFake images using Fourier analysis.

2- Training Support Vector Machines (SVMs) on these transformed features to distinguish different DeepFake generation methods.

3- The goal is to analyze the effectiveness of these approaches in detecting AI-generated faces and understanding the patterns behind them.

🚀 Stay tuned for updates and results!