Deepfake Detection Website 🎭

A Django-based web application for detecting deepfake videos using a hybrid CNN + LSTM deep learning model.
The system extracts frames from uploaded videos, processes them with a CNN to capture spatial features, and leverages an LSTM to model temporal dependencies for accurate detection.

---

🚀 Features

• Upload a video through a simple web interface.
• Preprocessing pipeline to extract frames from video.
• CNN extracts spatial (frame-level) features.
• LSTM models temporal sequence patterns.
• Returns deepfake probability and detection result.
• Django-powered website with clean UI.

---

🗂️ Project Structure

---

⚙️ Installation & Setup

1️⃣ Clone the repository

git clone https://github.com/Vikaskoppoju/deepfake.git
cd deepfake

2️⃣ Create a virtual environment

python -m venv .venv
source .venv/bin/activate   # On Linux/Mac
.venv\Scripts\activate      # On Windows

3️⃣ Install dependencies

pip install -r requirements.txt

4️⃣ Apply migrations

python manage.py migrate

6️⃣ Run the development server

python manage.py runserver

Now visit 👉 http://127.0.0.1:8000/ in your browser.

🧠 Model Details

• CNN extracts spatial features from individual frames.

• LSTM learns temporal dependencies across frame sequences.

• Model trained on deepfake video datasets (e.g., FaceForensics++ or DFDC).

• Outputs probability score and class label (Real / Fake).

📌 Usage

• Open the website in your browser.
• Upload a video file (.mp4, .avi, etc.).
• The system will:
  • Extract frames
  • Pass through CNN + LSTM model
  • Return whether the video is Deepfake or Authentic
• Results displayed on the UI.