Welcome to PhotoBooth!
Strike a pose and have fun with AI-powered styles.

PhotoBooth

PhotoBooth takes a picture of you from the webcam and generate an imaginary version of your picture, in different styles, using AI. It is a Python-based GUI application which applies AI-powered artistic styles using Comfy.ui. It includes a Raspberry Pi–based hotspot system (optional) for offline image sharing via a captive portal.

---

Features

• Capture images from a connected camera.
• Choose from multiple artistic styles.
• Apply styles using ComfyUI workflows. (styles are customizable)
• View, save, and share generated images.
• Local hotspot and captive portal for phone downloads without internet.

Photobooth has a log file where it records each times it generate an image. When an image is generated, it is saved for sharing via the Wifi hotspot, if activated. The image is deleted when Photobooth comes back to its welcome screen.

---

Hardware requirements

A gamer PC with a GPU is required. It works with a Nvidia 3050 but a bit slow (~20 seconds to generate an image). It has been tested on Windows and Linux. It should work decently with recent Macs. Connecting to a Raspberry Pi is optional and only required for sharing the generated pictures on a wifi hotspot run on the Raspberry Pi.

---

Project Structure

PhotoBooth/
├── comfy_classes/           # ComfyUI API integration
│   └── comfy_class_API.py
├── gui_classes/            # GUI logic and components (Pyside6)
│   ├── gui_manager/        # Managers for background, language, standby, threads, windows
│   ├── gui_object/         # GUI widgets: buttons, overlays, toolbox, etc.
│   └── gui_window/         # Window classes: main, base, sleep screen
├── gui_template/           # UI assets: icons, textures, gradients, sleep images
├── hotspot_classes/        # Hotspot and captive portal integration (Raspberry Pi)
├── language_file/          # Language translations (norway.json, sami.json, uk.json)
├── workflows/              # ComfyUI workflow definitions (default.json, clay.json, etc.)
├── constant.py             # Application constants and settings
├── prompts.py              # Style prompts for AI transformations
├── main.py                 # Main entry point for the application
├── requirements.txt        # Python dependencies
└── README.md               # Project documentation

---

Documentation

Configuration Guide

PhotoBooth uses two key files for configuration:

• prompts.py: Contains the dictionary dico_styles for all available artistic style prompts. You can customize or add new styles by editing this file. Each style entry defines how the AI should transform images in that style.
• constant.py: This file contains, in particular:
  • Interface appearance settings (colors, fonts, button styles, dimensions)
  • The messages displayed on the screen
  • Paths to working folders (input/output)
  • ComfyUI connection settings
  • The location of the workflows associated with each AI transformation
  • It is possible to fine-tune the appearance and behavior of the application by modifying this file.
  • Important note: Some constants present in this file are no longer used in the current version of the interface, but have been retained to ensure compatibility with older imports and avoid runtime errors.

Customization possible

Main elements you can edit in constant.py:

• Colors: The COLORS dictionary centralizes all the colors used.
• Text and button styles: Constants like BUTTON_STYLE, TITLE_LABEL_STYLE, DISPLAY_LABEL_STYLE, etc., precisely define the visual rendering.
• Size and layout: Ratios (DISPLAY_SIZE_RATIO, HUD_SIZE_RATIO) or margins (GRID_MARGIN_TOP, etc.) allow you to adjust the display on screens of various sizes.
• Custom AI prompt: The dico_styles variable allows you to associate each style with a unique prompt used in the build via ComfyUI.
• Timeout settings: SLEEP_TIMER_SECONDS, COUNTDOWN_START, etc.

Enable debug logs

There are two constants that control the amount of logs displayed in the console or saved in log files:

• DEBUG = False
• DEBUG_FULL = False

Set DEBUG = True to enable standard logs for monitoring normal execution. Set DEBUG_FULL = True to enable all logs, including frequent operations (e.g., refreshing the image), resulting in very verbose output.

Change connection addresses (ComfyUI, Flask, WebSocket)

If you have changed the network configuration or launched the services on different ports/addresses, update the following constants:

• WS_URL = "ws://127.0.0.1:8188/ws"

• HTTP_BASE_URL = "http://127.0.0.1:8188"

• HOTSPOT_URL = "https://192.168.10.2:5000/share"

• WS_URL: The URL of the WebSocket to receive build notifications from ComfyUI.

• HTTP_BASE_URL: HTTP URL to send requests to ComfyUI.

• HOTSPOT_URL: Local address of the Raspberry Pi Flask server (see dedicated section).

Disable hotspot image sharing

If you don't have a Raspberry Pi connected or you don't want to enable Wi-Fi sharing (captive portal), disable this option in constant.py:

• ShareByHotspot = False

After making changes, restart PhotoBooth to apply your new configuration.

---

Installation

Follow these steps to install and set up PhotoBooth.

1. Prerequisites

Hardware:

• Camera connected to your computer
• Optional: Raspberry Pi OS (for hotspot features, if using Raspberry Pi)

Software:

• comfyUI (see below for more details on installing ComfyUI),
• Git required by ComfyUI and photobooth, (install with default parameters)
• on Windows, ComfyUI requires Visual C++ and/or Visual Studio Community, you may need to install/re-install it (from Microsoft Store),
• Python 3.10, 3.11 or 3.12, also required by ComfyUI and photobooth,
• The installation process of ComfyUI and Photobooth will install the additional python packages needed for them to run,
• recommended: Visual Studio Code to modify the code (prompts and settings) and adapt it to your needs.

2. Clone the Repository

On Windows open a Terminal with by pressing the keys "Windows + r" and type in cmd and then press enter. In the Terminal, type cd Documents then:

git clone https://github.com/uitml/PhotoBooth.git
cd PhotoBooth

3. Install Python Dependencies

It is recommended to use a virtual environment, in the Photobooth folder run:

Windows (with Python 3.12)

python -p3.12 -m venv venv
venv\Scripts\activate.bat
pip install -r requirements.txt

Linux / Mac

python -m venv venv
source venv/bin/activate
pip install -r requirements.txt

4. Install Qt (Qt6) and System Libraries

On Windows, Qt is installed automatically via the installation of PySide6 in the pip install -r requirement.txt.

Linux

sudo apt update
sudo apt install qt6-base-dev
sudo apt install libxcb-cursor0 #may be needed

5. Configure ComfyUI

ComfyUI Installation & Model Setup

Comfy is assumed to be in the same folder as Photobooth. If this is not the case, be sure to set the correct path in constant.py.

Automated Setup using Scripts

To simplify the setup of ComfyUI's custom nodes and models, we have provided automation scripts for both Windows and Linux. These scripts will download and place all the required files for you.

• For Windows: Run the setup_comfyui.bat script by double-clicking it.
• For Linux/macOS: First, make the script executable by running chmod +x setup_comfyui.sh, then run it with ./setup_comfyui.sh.

If you encounter any issues with the scripts or prefer to set up your environment manually, please follow the detailed steps below.

---

1. Install ComfyUI

   • Follow instructions at: https://www.comfy.org/ or https://github.com/comfyanonymous/ComfyUI
   • On Windows ComfyUI installs by default in a folder ComfyUI in the Documents folder,
   • If you install ComfyUI from source using the Github repo you may go to your Documents folder and run these command lines (Linux):

     git clone https://github.com/comfyanonymous/ComfyUI.git
     cd ComfyUI
     python3 -m venv venv
     source venv/bin/activate
     pip install -r requirements.txt
     python main.py

2. Install Custom Nodes and Extensions

   • For enhanced features, search for and install the following ComfyUI custom nodes:

     • ComfyUI_IPAdapter_plus
     • comfyui_controlnet_aux
     • ComfyUI-Custom-Scripts
     • efficiency-nodes-comfyui
     • was-node-suite-comfyui

   • Automatic installation:

     • Use the ComfyUI Manager or extension installer (if available) to search for and install these nodes.

   • Manual installation:

     • Go to the ComfyUI/custom_nodes directory and run:

       git clone https://github.com/cubiq/ComfyUI_IPAdapter_plus.git
       git clone https://github.com/Fannovel16/comfyui_controlnet_aux.git
       git clone https://github.com/pythongosssss/ComfyUI-Custom-Scripts.git
       git clone https://github.com/jags111/efficiency-nodes-comfyui.git
       git clone https://github.com/ltdrdata/was-node-suite-comfyui.git

     • Restart ComfyUI after installing new nodes.

3. Download Required Models

   • ControlNet Models:
     • In ComfyUI/models/controlnet/ create the folder 1.5/ and place the 3 following safetensor files:
     • control_v11e_sd15_ip2p_fp16.safetensors, control_v11f1p_sd15_depth_fp16.safetensors, control_v11p_sd15_lineart_fp16.safetensors.
     • You can download them from:
       • HuggingFace and find the original documentation here: Depth and IP2P.
   • Stable Diffusion Checkpoints:
     • In ComfyUI/models/checkpoints/ place the Dreamshaper model, download the safetensor file: dreamshaper_8.safetensors from HuggingFace or Civitai (any SD 1.5 model can be used)
   • Upscale model
     • In ComfyUI/models/upscale_models/ place the file RealESRGAN_x2.pth, you can find it on HuggingFace
   • Other Models:
     • For IP-Adapter, follow the procedure in the ComfyUI_IPAdapter_plus documentation. Make sure you have in your folder ComfyUI/models/clip_vision and place the following files:
       • CLIP-ViT-H-14-laion2B-s32B-b79K.safetensors
       • clip-vit-large-patch14.safetensors
     • In folder ComfyUI/models/ipadapter (create folder if it does not exist):
       • ip-adapter-faceid-plusv2_sd15.bin
       • ip-adapter-faceid-portrait-v11_sd15.bin
       • ip-adapter-faceid_sd15.bin
       • ip-adapter-plus-face_sd15.safetensors
       • ip-adapter_sd15.safetensors
     • In folder ComfyUI/models/loras:
       • ip-adapter-faceid-plusv2_sd15_lora.safetensors
       • ip-adapter-faceid_sd15_lora.safetensors

You may need the following additional AI/FaceID dependencies, if you have the error "No module named 'insightface'" when you try to run one of the Photobooth workflows (see below 'Test ComfyUI'),

For Windows, in the Terminal, go to the Comfy folder and run:

.venv\Scripts\activate.bat
pip3 install insightface
pip3 install onnxruntime

for Linux inside th evirtual environment of ComfyUI:

pip install insightface
pip install onnxruntime
pip install onnxruntime-gpu

4. Test ComfyUI Start ComfyUI and verify that models in the workflows directory of PhotoBooth can be loaded correctly in ComfyUI and run on a test image (use a test image with a face present in it).

6. Hotspot & Captive Portal Setup (Raspberry Pi)

This is optional and only necessary if you want to use the Rasberry Pi to share images using Wifi. Follow the steps in the PDF to set up the hotspot and captive portal. Place configuration files in hotspot_classes/in_py/configuration_files/ as described.

7. Run PhotoBooth

Open a Terminal or use the one opened previously, in the PhotoBooth folder, run:

Linux/macOS

python main.py

Windows

venv\Scripts\activate.bat
python main.py

You can use a .bat script to automate launching PhotoBooth and ComfyUI (change the paths to your correct comfy and photobooth paths):

@echo off
del /f /q "C:\AI Demos\PhotoBooth\app.log"
start "" "C:\Users\USERNAME\AppData\Local\Programs\@comfyorgcomfyui-electron\ComfyUI.exe"
cd /d "C:\AI Demos\PhotoBooth"
python .\main.py
pause

This script does the following:

• Deletes the previous log file.
• Launches the ComfyUI executable (via ComfyUI Electron).
• Starts the PhotoBooth app.
• Leaves the window open (useful for checking for any error messages).

Note: Adapt this .bat file to your own configuration, including the paths to ComfyUI.exe and the PhotoBooth folder.

Communication between PhotoBooth and ComfyUI

Photobooth assumes that ComfyUI is running and accessible at the local url HTTP_BASE_URL = "http://127.0.0.1:8188". If it is serving another url, you can modify the constant.py to point to the correct one. More generally, the communication between the PhotoBooth app and ComfyUI is set in the constant.py file (see config details). Check it if PhotoBooth does not work after ComfyUI is set. Restart PhotoBooth after any modification of the configuration file.

8. Quit Photobooth

Just press the keys alt and F4 to close the window. The Photobooth window is configured to be always in front.

---

Credits

Developed as part of the Machine Learning Group – UiT Tromsø demonstration projects.
Full list of contributors and acknowledgements are included in the PDF documentation CR_Installation_Photobooth_2025_V3_en.pdf.