Extensions/COMFYUI-MODELINSTALLER
ComfyUI Extension

COMFYUI-MODELINSTALLER

Comprehensive ComfyUI extension pack featuring one-click model installation, real-time download progress, video processing utilities for eye stabilization and character face swapping, and system utilities.

By Pis7i·Created 9 months ago·Updated 8 months ago· 0
Pis7i/COMFYUI-MODELINSTALLER
Nodes
On cloudLocal install
Stars0
Updated8 months ago
Readme

ComfyUI Model Installer + PMA Utils

A comprehensive ComfyUI extension pack featuring model management, video processing utilities, and custom nodes.

Features

Model Management

  • One-click model installation from the ComfyUI interface
  • Real-time download progress in a dedicated window
  • Automatically organizes models into their correct folders
  • Resumes interrupted downloads
  • Skips already downloaded models

Video Processing Nodes

  • Eye Stabilizer: Fixes eye glitching in video motion transfer workflows (Wan2.2, etc.)
  • Character Swap: ControlNet-based character face swapping
  • Temporal smoothing and enhancement tools

System Utilities

  • Automatic shutdown monitoring with queue detection
  • Activity tracking to prevent premature shutdowns
  • ComfyUI session management tools

Installation

  1. Navigate to your ComfyUI custom_nodes directory:
cd ComfyUI/custom_nodes
  1. Clone or copy this extension:
git clone https://github.com/yourusername/ComfyUI-ModelInstaller.git

Or manually copy the ComfyUI-ModelInstaller folder to your custom_nodes directory.

  1. Restart ComfyUI

Usage

  1. Launch ComfyUI
  2. Click on Model ManagerInstall Models in the top menu bar
  3. A dialog window will open
  4. Click Start Download to begin downloading all models
  5. Monitor the progress in real-time
  6. Close the window when complete

Requirements

Core Requirements

  • wget must be installed on your system
    • Linux/Mac: Usually pre-installed
    • Windows: Install via chocolatey: choco install wget

For Eye Stabilizer Node

  • MediaPipe (for face detection):
    pip install mediapipe==0.10.21
    
    • If not installed, node falls back to basic mode without face tracking

Models Included

The extension downloads the following models:

  1. ControlNet Union SDXL - Advanced control model
  2. GFPGAN Face Restore - Face restoration model
  3. CLIP Vision SDXL - CLIP vision encoder
  4. GonzalomoXL Checkpoint - Main checkpoint model
  5. IP-Adapter FaceID Plus v2 - Face identity preservation
  6. SwinIR 4x Upscaler - Image upscaling model
  7. EpiC Negative Embedding - Negative prompt embedding
  8. Deep Negative Embedding - Additional negative embedding

Adding More Models

To add more models to the download list, edit model_downloader.py:

MODELS = [
    {
        "url": "https://example.com/model.safetensors",
        "path": "folder/filename.safetensors",
        "name": "Display Name"
    },
    # Add more models here
]

Troubleshooting

wget not found error:

  • Install wget on your system
  • Linux: sudo apt-get install wget or sudo yum install wget
  • Mac: brew install wget
  • Windows: choco install wget

Downloads fail:

  • Check your internet connection
  • Verify the URLs are still valid
  • Ensure you have sufficient disk space

Custom Nodes Documentation

Eye Stabilizer

Fixes eye glitching, jittering, and unnatural blinking in video motion transfer workflows.

Full Documentation: EYE_STABILIZER_README.md
Setup Guide: SETUP_EYE_STABILIZER.md
Technical Details: EYE_STABILIZER_IMPLEMENTATION.md

Features:

  • Dense eye landmark detection (478 facial points via MediaPipe)
  • Temporal smoothing using Kalman filtering
  • Blink detection and synthesis
  • Eye region enhancement
  • Debug visualization

Quick Start:

Video → RMBG → [Eye Stabilizer] → DWPose/Depth → WanVaceToVideo

Test Installation:

python test_eye_stabilizer.py

Character Swap

ControlNet-based character face swapping for pose/scene transfer.

Status: Basic implementation (preprocessors only)
Planned: Full IP-Adapter FaceID integration

Project Structure

ComfyUI-ModelInstaller/
├── __init__.py                          # Main entry point
├── model_downloader.py                  # Model download system
├── shutdown_monitor.py                  # Auto-shutdown utilities
├── character_swap_node.py              # Character swap node
├── eye_stabilizer_node.py              # Eye stabilizer node (NEW)
├── test_eye_stabilizer.py              # Test suite (NEW)
├── README.md                            # This file
├── EYE_STABILIZER_README.md            # Eye stabilizer docs
├── SETUP_EYE_STABILIZER.md             # Setup guide
├── EYE_STABILIZER_IMPLEMENTATION.md    # Technical docs
└── js/
    └── model_installer.js               # Frontend UI

License

MIT