MagicQuill Intelligent Interactive Image Editing System

By /

Project name: MagicQuill
Project function: AI image editing
Project Description: An intelligent interactive image editing system that enables accurate image editing through a user-friendly interface and AI-driven suggestions.
Easily edit images through simple operations such as drag-and-drop and drawing, suitable for users of all levels, including beginners.
Features include local precise editing, smart prompts and a variety of brush tools, allowing users to add or delete image elements according to their needs.

project overview

MagicQuill by Ant Technology Research Institute developed Intelligent interactive image editing system, aims to achieve efficient and accurate image editing. The project combines a user-friendly interface with powerful AI capabilities, allowing users to complete complex editing tasks with minimal input, including:

  • ✏️ insert elements
  • 🚫 delete objects
  • 🎨 change the color

Project address:MagicQuill GitHub repository

🌟 Focus on core functions

  1. intuitive user interface
    Provides a simple and powerful interactive interface that allows users to easily operate by painting with brushes, clicking, etc.

  2. AI-driven smart advice
    Adopt a multimodal large language model (MLLM) to predict editing intentions in real time without the need to manually enter complex commands.

  3. Precise local editing
    Relying on the diffusion model and dual-branch plug-in module, high-quality local image modification is achieved with accurate details.

🚀Update log

  • November 21, 2024

    • Add save button
    • Fix path errors on Windows
    • Provides installation scripts for Windows and Linux environments

  • December 6, 2024

    • Add automatic saving function
    • Support resolution adjustment

  • January 2, 2025

    • Release Docker support
    • Allowing users to build images in an isolated environment

  • February 27, 2025

    • 🎉 MagicQuill is officially accepted by CVPR 2025!

Hardware requirements

configuration items minimum requirements
operating system Windows / Linux
processor multicore CPU
memory At least 16GB RAM
graphics card NVIDIA GPU (6GB memory)

🏁Installation and use

  1. environment settings
    Select the installation script based on the operating system:

    
# Windows
windows_setup.bat

# Linux
linux_setup.sh

GitHub：https://github.com/magic-quill/magicquill

  1. Install dependencies
    Make sure Docker is installed and use Dockerfile to build the image:
docker build -t magicquill .
  1. launch an application
    Run the following command and access it through the Gradio interface:
python gradio_run.py

Research citation

If you use MagicQuill in your research, please quote the following papers:

Zichen Liu, Yue Yu, Hao Ouyang, Qiuyu Wang, Ka Leong Cheng, Wen Wang, Zhiheng Liu, Qifeng Chen, Yujun Shen,
MagicQuill: An Intelligent Interactive Image Editing System, CVPR 2025.

Paper link:arXiv

GitHub：https://github.com/magic-quill/magicquill
Oil tubing:

Scroll to Top