Easy Danbooru Tagger for ComfyUI with Clothing & Pose Separation (Image Captioning) — v2.0

This ComfyUI workflow is for captioning and tagging a loaded image, separates clothing and pose information, and produces image-captioning or tagging text.

It adds one filter for character attributes, such as 1girl, 2boys, and skin tone, and another for background and environment tags.

It places action tags such as skirt grab in Poses/Actions instead of also listing them under Clothing.

Input: an image loaded in ComfyUI.

Output: image-captioning or tagging text with clothing and pose information separated.

Load your image in ComfyUI and press Run Workflow.

The workflow requires these node packs: cg-use-everywhere, ComfyUI-JPS-Nodes, ComfyUI-Easy-Use, ComfyUI-Custom-Scripts, and ComfyUI-WD14-Tagger.

It also lists was-ns and comfyui_tag_fillter as node pack dependencies.

Version v2.0 adds character-attribute and background/environment filters.

The tagger settings can be customized to your preference.

Before setup, verify the package names for was-ns and comfyui_tag_fillter.

Vorschlag · nicht geprüft

No license is listed; verify usage terms before redistributing the workflow.

Vorschlag · nicht geprüft