Selected work / 01Ongoing practice

ComfyUI

Node-based workflows for image editing, generative video and character-to-3D creation.

Disciplines: ComfyUI / Generative Video / LoRA / Workflow Integration
Status: Ongoing practice

Inside the workflow

A look inside my ComfyUI setup: a source image passes through BiRefNet background removal, a cutout mask, crop and canvas controls, then preview and export.

ComfyUI workflow recording — source image, cutout, crop, preview and export.

Portrait workflow & review

Two recordings from my avatar workflow: the ComfyUI node graph and batch queue, followed by a review of portrait variations.

ComfyUI avatar workflow — node graph and batch queue

Portrait review — browsing generated variations

Portrait examples

Generated portrait variations across poses, clothing and settings.

Generated portrait example: Front-facing portrait — neutral background — Front-facing portrait — neutral background

Side-profile portrait

Generated portrait example: Three-quarter studio portrait — Three-quarter studio portrait

Generated portrait example: Outdoor portrait — city street — Outdoor portrait — city street

Generated portrait example: Office portrait — Office portrait