A full-stack AI-powered web application that generates thumbnails and continuously improves them based on user feedback (like / dislike) until the user is satisfied.
- User requests a thumbnail
- AI generates a thumbnail using Hugging Face
- User clicks Like or Dislike
- If Liked → thumbnail is accepted
- If Disliked → AI regenerates an improved thumbnail
- Process repeats until the user likes the result
- AI-based thumbnail generation
- Like / Dislike feedback loop
- Automatic thumbnail improvement
- Hugging Face inference integration
- Modern and responsive UI
- React + Vite
- TypeScript
- Node.js + Express
- Hugging Face Inference API
- Git & GitHub
📁 Project Structure
├── public
├── src
│ ├── components # UI components
│ ├── pages # Application pages
│ ├── context # Global state management
│ ├── services # API calls
│ └── utils # Helper functions
├── server
│ ├── controllers # Request handling logic
│ ├── routes # API routes
│ ├── services # AI & business logic
│ └── config # Environment & setup
├── package.json
└── README.md
This is the main landing page of the application where users can start generating AI-based thumbnails.
This thumbnail is generated by the AI based on the user’s input before any feedback is provided.
When the user clicks Dislike, the system uses the feedback to regenerate and improve the thumbnail automatically.
