ComfyUI Extension: Cozy Human Parser

Cozy Human Parser Introduction

The human-parser-comfyui-node extension is a powerful tool designed to help AI artists generate detailed masks for specific body parts and clothing items in images. This extension is particularly useful for those working in fashion AI, digital art, and other creative fields where precise human parsing is essential. By using this extension, you can easily identify and isolate various elements within an image, such as hats, hair, gloves, and more, making it easier to manipulate and enhance your artwork.

How Cozy Human Parser Works

The human-parser-comfyui-node extension works by leveraging pre-trained models to analyze images and generate masks for different body parts and clothing items. These models have been trained on large datasets, allowing them to accurately identify and segment various elements within an image. The extension can run on both CPU and CUDA, making it accessible to a wide range of users.

Think of it as a highly skilled digital artist who can quickly and accurately outline different parts of a human figure in an image. Once these parts are outlined, you can easily apply different effects, colors, or modifications to each part, giving you greater control over your creative process.

Cozy Human Parser Features

The human-parser-comfyui-node extension comes with several key features that make it a valuable tool for AI artists:

• Fast and Efficient: The extension is designed to be fast and lightweight, ensuring that it doesn't consume too much VRAM and can run smoothly on most systems.
• Versatile Mask Generation: It can generate masks for a wide range of body parts and clothing items, allowing for detailed and precise editing.
• CPU and CUDA Support: Whether you have a powerful GPU or are working on a CPU-only system, the extension can adapt to your hardware capabilities.

Cozy Human Parser Models

The extension supports three different models, each trained on a specific dataset. These models offer different levels of detail and are suited for various applications:

LIP Parser

• Dataset: LIP (Look into Person)
• Images: 50,000+
• Focus: Complicated real scenarios
• Categories: 20 (e.g., Background, Hat, Hair, Glove, Sunglasses, Upper-clothes, Dress, Coat, Socks, Pants, Jumpsuits, Scarf, Skirt, Face, Left-arm, Right-arm, Left-leg, Right-leg, Left-shoe, Right-shoe)
• mIoU: 59.36%
• Example:

ATR Parser

• Dataset: ATR
• Images: 17,000+
• Focus: Fashion AI
• Categories: 18 (e.g., Background, Hat, Hair, Sunglasses, Upper-clothes, Skirt, Pants, Dress, Belt, Left-shoe, Right-shoe, Face, Left-leg, Right-leg, Left-arm, Right-arm, Bag, Scarf)
• mIoU: 82.29%
• Example:

Pascal Parser

• Dataset: Pascal Person Part
• Images: 3,000+
• Focus: Body parts segmentation
• Categories: 7 (e.g., Background, Head, Torso, Upper Arms, Lower Arms, Upper Legs, Lower Legs)
• mIoU: 71.46%
• Example:

Troubleshooting Cozy Human Parser

Here are some common issues you might encounter while using the extension and how to solve them:

Ninja is required to load C++ extensions

If Windows can't find the "ninja.exe" file, you need to locate it and add its path to the system PATH. For more details, see .

NK1104: cannot open file 'python311.lib'

If Windows can't locate the python311.lib library, search for it on your system and add its parent directory to the "LIB" environment variable. For more details, see .

Command '['where', 'cl']' returned non-zero exit status 1

If Windows can't locate "cl.exe", ensure you've installed all necessary components and start ComfyUI from the developer command prompt. For more details, see .

error: first parameter of allocation function must be of type "size_t"

Make sure you're running the "x64 Native Tools Command Prompt" instead of the x86 one. Type "x64" in the start menu to locate it.

Learn More about Cozy Human Parser

To learn more about the human-parser-comfyui-node extension, you can explore the following resources:

• These resources provide additional information, tutorials, and community support to help you get the most out of the extension.