This repository provides a ComfyUI node for the IDM-VTON virtual try-on model, enabling users to generate realistic virtual try-on images. It targets users of ComfyUI interested in AI-powered fashion and image manipulation.

How It Works

The node integrates the IDM-VTON model into the ComfyUI workflow. It leverages ComfyUI's Segment Anything node for mask generation and ControlNet Auxiliary Preprocessors for DensePose estimation, streamlining the virtual try-on process within a familiar node-based interface.

Quick Start & Requirements

• Install: Via ComfyUI Manager or by cloning the repo into custom_nodes and running python install.py.
• Prerequisites: GPU with at least 16GB VRAM.
• Models: Weights from yisol/IDM-VTON will be downloaded automatically.
• Links: ComfyUI Manager

Highlighted Details

• ComfyUI adaptation of the IDM-VTON virtual try-on model.
• Utilizes ComfyUI Segment Anything for mask generation.
• Employs ComfyUI's ControlNet Auxiliary Preprocessors for DensePose estimation.

Maintenance & Community

• Project is maintained by TemryL.
• Encourages contributions via pull requests.

Licensing & Compatibility

• Original IDM-VTON source code is under CC BY-NC-SA 4.0 license.
• This license may impose restrictions on commercial use and redistribution.

Limitations & Caveats

Requires a high-end GPU with a minimum of 16GB VRAM, which may be a barrier for users with less powerful hardware.