1. Workflow Overview

Purpose:
Generates multiple facial expressions from a single portrait using ExpressionEditor nodes, ideal for actor auditions or character design.

Key Features:

• Produces 6 expressions (e.g., smile, anger, surprise) from one input image

• Supports localized edits (eyes/mouth/full face)

• Outputs high-quality expression sequences

2. Key Components & Installation

Required Nodes:

• ExpressionEditor: Core facial control (install FaceChain or Expression-Control plugin)

• LoadImage: Loads input portrait (JPG/PNG)

• SaveImage: Saves results (default: ComfyUI/output)

Dependencies:

• Input image must be front-facing and clear (resolution ≥512x512)

• Plugin: Install Expression Editor via ComfyUI Manager

3. Workflow Structure

Group 1: Input

• Node: LoadImage

• Input: Single portrait (e.g., actor_photo.jpg)

• Output: Distributes image to 6 ExpressionEditor nodes

Group 2: Expression Editing

• Nodes: 6x ExpressionEditor (each controls a unique expression)

  • Parameters:

    • OnlyEyes: Adjusts eyes (e.g., dilated pupils + raised brows)

    • OnlyExpression: Full-face edits (e.g., downturned mouth + frown)

  • Controls:

    • Brow height/angle

    • Eye openness/pupil size

    • Mouth shape/teeth visibility

Group 3: Output

• Node: SaveImage

• Result: Saves expression images (individual or merged)

4. Inputs & Outputs

Input Parameters:

• Required: Clear front-facing portrait

• Optional: Expression intensity (default 1.7), local locks (e.g., eyes only)

Output:

• 6 PNG images with different expressions

5. Notes

• Image Quality: Low resolution may distort facial features

• Naturalness: Keep intensity ≤2.0 for realistic results

• Performance: CPU-dependent (GPU acceleration minimal)