Truth Fairy empowers you to spot misinformation as you browse. Whether it’s a suspicious text snippet or a manipulated image, our extension brings the truth to light—fast and fun!
- Instant Verification ✅/❌: Simply highlight content. If it’s false, it turns red; if true, it briefly glows green with extra info.
- Image Analysis 👁️: Our enhanced deep learning model, powered by explainable AI (using Grad-CAM), detects deepfakes and highlights manipulated regions.
- Community Engagement 💬: Join real-time discussions and view fact-checked comments via our built-in panel.
- Misinformation Scoring 📊: Each statement is scored to indicate its reliability.
Set up your development environment with the following commands:
# Create and activate the virtual environment
python -m venv env
source env/bin/activate
# Install dependencies
pip install -r requirements.txt
# Install backend dependencies
cd extension/backend
npm install
cd ../..Follow these steps to launch the application:
- Create a .env file with the following content
# Create a .env file and add the OPEN_AI and BRAVE_API keys. If you need the keys, please contact [email protected].
OPENAI_KEY = "ACTUAL OPENAI API KEY"
BRAVE_API_KEY = "ACTUAL BRAVE API KEY"- Start the Backend Server:
cd extension/backend
node index.js
#It'll prompt you for firebase_credentials.json, please contact [email protected], then rerun the node command.- Start the Deepfake Detection Server:
cd ../..
cd models/deepfake_detection
# Place the pretrained weights in the 'pretrained_model' folder (see link below)
pip install .
cd ..
python app.py- Open Google Chrome and go to Manage Extensions.
- Click Load Unpacked and select the
frontendfolder. - Enjoy safer browsing and join the conversation!
Download the pretrained weights here:
Pretrained Weights on Google Drive
Feel free to contribute and help us build a more truthful online community!