2025-03-26 21:57:50 +01:00
2025-03-26 21:57:50 +01:00
2025-03-26 21:57:50 +01:00
2025-03-26 21:57:50 +01:00
2025-03-26 21:57:50 +01:00

✨ Avatar Generation Experience

A custom node for ComfyUI that transforms user photos into stylized character avatars.

🔍 Project Overview

Avatar Generation Experience is an interactive application developed as a custom node for ComfyUI. The project creates an engaging user experience that transforms ordinary photos into stylized avatars across different fictional universes and themes.

🚀 Key Features

  • Themed Character Groups: Includes Marvel, DC, Disney, Pixar, historical figures, and film characters
  • Real-time Photo Processing: Captures and transforms user photos on the fly
  • Interactive UI: Intuitive interface for character selection and avatar generation
  • QR Code Scanning: Scan user QR code to start the experience
  • QR Code Sharing: Generated avatars can be shared via QR codes
  • Screensaver Gallery: Displays previously generated avatars when the system is idle
  • Google Drive Integration: Cloud storage and sharing capabilities
  • Simplified Controls: The entire experience can be navigated using just a three-button joystick, designed for exhibition on large vertical screens at trade shows and events

🛠️ Technical Implementation

  • Custom ComfyUI Node: Extends ComfyUI's functionality with specialized image processing
  • WebSocket Communication: Real-time updates during the generation process
  • IndexedDB Storage: Local caching of generated images
  • REST API Integration: Communicates with backend services for image processing
  • Responsive Design: Works across different device sizes

📸 Screenshots

Application Interface Generated Avatar
Main interface showing Themed Character Groups selection Example of customization options available for avatar generation

💻 Technologies Used

  • JavaScript
  • WebSockets
  • IndexedDB
  • Google Drive API
  • ComfyUI Framework
  • Python (for QR code scan)
S
Description
No description provided
Readme
3.9 MiB
Languages
JavaScript 71.2%
CSS 19.6%
HTML 5.8%
Python 3.4%