Multi-Person Changer: A 4-Stage ComfyUI Workflow for Character Transformation
Any-Scar765 · reddit · 2026-07-22
This post shares a custom ComfyUI workflow called Multi-Person Changer, providing a complete 4-stage AI pipeline for character transformation.
Core Features & Pipeline:
- Strip: Removes the background via RMBG and strips the character down to a clean base body.
- Pose Transfer: Extracts the pose/action from a reference image and applies it to the target character.
- Product Outfit: Processes the reference outfit image into a clean, product-style showcase.
- Outfit Attachment: Composites the processed outfit onto the transformed character at high resolution.
Technical Details:
- Supports video processing using PySceneDetect for automatic scene boundary detection and frame-by-frame application.
- Uses KJNodes Set/Get nodes to pass data without visual spaghetti.
- Enables caching via a fixed Seed, instantly skipping redundant computations when re-queueing without parameter changes.
More from Multimodal
- Midjourney v8.2 preview shows a neon-flower portrait aesthetic — chrisfirst · 2026-07-22
- Midjourney v8.2 preview teases a neon-flower portrait style — AIFlow_ML · 2026-07-22
- Can image-to-3D make realistic 3D-printable human figurines without losing facial identity? — LoudPoem9492 · 2026-07-22
- ZibraAI demos ZibraGDS live at SIGGRAPH with graphics-heavy stage footage — MonaJalal_ · 2026-07-22
- Fable users demo a procedural dog simulation built with three.js agent skills — willdepue · 2026-07-22
- Scenario Launches Cartwheel Character Rigging: Single Image to Fully Rigged 3D Model — andrew_n_carr · 2026-07-22