Deepfake Detection Web App — API Documentation

Welcome to the API documentation for the Deepfake Detection Web App. This web service allows users to sign up, log in, and upload video files to detect possible deepfake manipulations using basic blur artifact analysis.

This documentation is structured to help developers integrate and use the backend API effectively. Each route is explained with request types, expected data, and response examples in JSON or HTML format.

---

🌍 Base URL

http://localhost:5000/

All routes mentioned below are based on this default development URL. Change it accordingly if deployed in production.

---

📑 Endpoints Overview

Method	Endpoint	Description
GET	/	Renders the login page
POST	/login	Authenticates a user and starts a session
GET	/signup	Renders the signup page
POST	/signup	Registers a new user
GET	/home	Renders user home/dashboard (after login)
POST	/upload	Uploads a video for deepfake analysis
GET	/logout	Logs the user out and ends the session
GET	/ststic/<filename>	Serves static files from the ststic/ directory

---

📌 Expected Difference (based on research and experiments):

Type Avg Blue Intensity

Real ~110–130

Deepfake ~90–110

📉 Deepfakes often show lower blue levels and less blue contrast overall, especially in areas where the fake face is blended.

📝 Signup - POST /signup

Registers a new user with Gmail-only email validation.

Request (form-data)

{
  "username": "preeti",
  "email": "preeti@gmail.com",
  "password": "SecurePassword"
}

Success Response (HTML - signup_success.html)

<h3>Signup successful, welcome preeti!</h3>

Failure Response (HTML - signup_error.html)

<h3>Only Gmail addresses are allowed!</h3>

or

<h3>Username or Gmail already exists!</h3>

---

🔐 Login - POST /login

Authenticates the user and stores a session.

Request (form-data)

{
  "username": "preeti",
  "password": "SecurePassword"
}

Success

• Redirects to /home.

Failure (HTML - login_error.html)

<h3>Invalid credentials</h3>

---

🏠 Home - GET /home

Displays the dashboard after a successful login.

• 🔒 Requires user session (session['user']).
• Redirects to / if not logged in.

---

📤 Upload Video - POST /upload

Allows the user to upload a video file for blur-based deepfake detection.

Request (multipart/form-data)

Field	Type	Description
video	file	The video file to test

Response (HTML - result.html)

🛑 Deepfake Detected!

or

✅ Video Looks Real.

---

🚪 Logout - GET /logout

Logs out the current user by clearing the session.

Response

• Redirects to /.

---

🗂️ Serve Static Files - GET /ststic/<filename>

Serves CSS, JS, images, etc., from the ststic/ directory.

⚠️ You have a typo in the folder name (ststic instead of static). Either fix it or document it properly for consistency.

---

Let me know when you're ready, and I’ll proceed with a detailed security audit covering:

• 🔐 Session management
• 🛡️ SQL injection
• 🗂️ File upload validation
• ❌ XSS and CSRF
• 📁 Directory traversal
• 🔄 Misc recommendations

Would you like me to start the security audit now?