Skip to content

Latest commit

 

History

History
87 lines (62 loc) · 3.29 KB

README.md

File metadata and controls

87 lines (62 loc) · 3.29 KB
Raw

Deepfake Detector App

ScreenShots

Streamlit - Google Chrome 16-01-2024 21_30_24 Streamlit - Google Chrome 16-01-2024 21_30_40 Streamlit - Google Chrome 16-01-2024 21_50_22 Streamlit - Google Chrome 16-01-2024 21_33_16

Table of Contents

Introduction

Welcome to the Deepfake Detector App repository! This Streamlit app is designed to detect deepfake content in images and videos using state-of-the-art models. It provides a user-friendly interface for uploading files and obtaining deepfake predictions with adjustable parameters.

Recommended parameters:

  • Model : EfficientNetAutoAttB4
  • Dataset: DFDC

Features

  • File Type Selection: Choose between uploading an image or a video for deepfake detection.
  • Model Selection: Select from various deepfake detection models, such as EfficientNetB4, EfficientNetB4ST, EfficientNetAutoAttB4, etc.
  • Dataset Option: Choose the dataset (DFDC or FFPP) used to train the deepfake detection model.
  • Adjustable Threshold: Set a threshold for deepfake probability to control sensitivity.
  • Video Frame Selection: If analyzing a video, choose the number of frames to process.
  • Detailed Results: Get detailed results with probabilities and visual cues indicating the likelihood of deepfake content.
  • Project Information: Display additional information about the project, such as credits, links to GitHub, and collaborators.

Installation

  1. Clone the repository:
git clone https://github.com/Sneh-T-Shah/deepfake-detection.git
cd deepfake-detection
  1. Install the required dependencies:
pip install -r requirements.txt

Usage

Run the Streamlit app:

streamlit run app.py

Visit the provided local URL to access the app in your browser.

File Structure

  • app.py : Main Streamlit application script.
  • api.py : Contains functions for processing images and videos using deepfake detection models.
  • uploads/ : Folder to store uploaded files.
  • requirements.txt : List of Python dependencies.

Dependencies

Find the dependencies here: https://github.com/Sneh-T-Shah/deepfake-detection/blob/main/requirements.txt

Contributing

We welcome contributions! If you'd like to contribute to this project.

Acknowledgments

Web app for this project is made by Sneh Shah and Pankil Soni.

The original source for the deep-learning models is on the github reopsitory https://github.com/polimi-ispl/icpr2020dfdc

Contact

For any query or feedback, please contact: