ivaaak / ArchAI

ArchAI - React / Express / StableDiffusion / ControlNet- generate architectural sketches and drawings (WIP)

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

ArchAI - React / Express / StableDiffusion - AI Image generation for architects:

A Web App built with React as a Frontend and Express as a Backend. It uses StableDiffusionXL and ControlNet for Image generation and modification. Data stores: MongoDB, Cloudinary (aswell as an on-server uploads folder*).

Screenshots:

Getting Started:

You need the following API keys to add to a .env file in the backend folder:

ATLAS_URI= (mongoDB connection string)
REPLICATE_API_TOKEN= (StableDiffusion / Replicate API Key)
CLOUD_NAME= (cloudinary name / api key / secret)
CLOUDINARY_API_KEY=
CLOUDINARY_API_SECRET=

You can run the below commands from the ArchAI directory and start the project:

npm i
npm start

This installs and starts both the FE and BE using the npm tool 'concurrently'. Or you can run the commands separately in the frontend / backend folders to have them running in separate instances/terminals.

Built With:

  • [] React (Vite, Typescript)
  • [] Express API
  • [] StableDiffusion
  • [] ControlNet
  • [] Auth0
  • [] Axios
  • [] Cloudinary
  • [] MongoDB
  • [] Stripe

Features / Image generation modes:

  • Text to Image
  • Image to Image
  • Sketch (canvas drawing) to Image
  • Image Infill
  • Auth0 Auth and User Management
  • Collections
  • Sharing Prompts / Images generated

Not implemented yet / In Progress:

  • Image In-Painting / Infill (only re-generate part of an Image)
  • Support for Midjourney and a switch between it and StableDiffusion
  • Custom trained LLMs for specific image modes - photorealistic / pencil sketch / 3d render
  • LLM Fine-Tuning
  • 3D Models - generation and CAD integration
  • Discover - Analytics and Feedback to provide users with analytics on their sketching habits and performance.

About

ArchAI - React / Express / StableDiffusion / ControlNet- generate architectural sketches and drawings (WIP)


Languages

Language:TypeScript 82.6%Language:CSS 16.4%Language:JavaScript 0.6%Language:HTML 0.3%Language:Dockerfile 0.2%