rudnibh/Photobooth ? reverse-engineered prompt

Build me a web based AI photo booth that lets someone use their webcam, take a picture, and turn it into a stylized portrait with AI.

I want the flow to feel simple and fun, like a kiosk app. The user opens the page, sees their camera, snaps a photo, picks from a few art styles, and waits for the generated result. The app should send the image to a Stable Diffusion backend with pose and face detail preserved as much as possible, then show the finished image and save it so it can be viewed or downloaded later.

Please make it work with a separate image server for uploads, and set it up so I can easily point it at my own local AI server URLs. If you need to check current setup details, look up the latest docs online.