An AI-powered web application that detects whether a video is Real or Fake using deep learning techniques. This system analyzes video frames, extracts facial features, and performs classification using a trained model.
DeepFake content is becoming increasingly realistic and dangerous. This project aims to provide a reliable solution to identify manipulated videos using a trained deep learning model.
The application allows users to upload videos and get predictions along with confidence scores.
- 🎥 Upload and analyze video files
- 🧍 Face extraction from video frames
- 🤖 Deep learning-based classification (Real vs Fake)
- 📊 Confidence score for predictions
- ⚡ Fast and interactive UI using Streamlit
- Frontend/UI: Streamlit
- Backend: Python
- ML Framework: TensorFlow / Keras
- Computer Vision: OpenCV
- Other Libraries: NumPy
Deep_Fake/
│
├── app.py
├── requirements.txt
├── README.md
├── .gitignore
└── model_final.keras/ (not included in repo)
git clone https://github.com/Dinesh-MDT/DeepFakeDetection.git
cd DeepFakeDetection
pip install -r requirements.txt
Download the model from the link below:
👉 https://drive.google.com/file/d/1acCBpAedUVP0WkI_0j069dBxk-YIF_NW/view?usp=sharing
After downloading:
- Extract the ZIP file
- Place the folder inside the project directory
DeepFakeDetection/
└── model_final.keras/
streamlit run app.py
Add screenshots of your application here
Example:
- Upload screen
- Prediction result screen
- The model file is not included due to size limitations
- Make sure the model is placed correctly before running the app
- If the model is missing, the application will not work
- 🔄 Real-time video detection
- 🌐 Deploy as a web application (Streamlit Cloud / AWS)
- 📈 Improve model accuracy with larger datasets
- 🧪 Add evaluation metrics and visualizations
Contributions are welcome. Feel free to fork the repository and submit pull requests.
This project is for educational purposes.
If you found this project useful, consider giving it a star ⭐