OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Image & Video AI

Comfyui H3 Facerefine

ComfyUI-H3-FaceRefine is a MIT-licensed set of custom nodes for refining small faces in MiniMax H3 video. It detects and tracks faces frame by frame, crops a selected face so it fills a canvas for H3 regeneration, and composites the result back into the original frames.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Comfyui H3 Facerefine — OpenViST project profile in Image & Video AI.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it in a ComfyUI MiniMax H3 workflow when a subject’s face occupies a small part of the video frame and needs frame-by-frame refinement.

Who it is for

  • ComfyUI users working with MiniMax H3 video who want to refine faces that appear small in the frame.
  • Workflow builders who need automatic or per-shot manual face selection.

Documented uses

  • Track a selected face across video frames.
  • Regenerate a larger crop of the face with H3.
  • Composite refined face imagery back into the original video.
  • Choose a subject automatically by ranking or manually by shot.
Before you choose

Requirements & limits

The project provides a public GitHub repository under the MIT license.

  • Requires ComfyUI with MiniMax H3 support and a separately supplied face detector.
  • Automatic hard-cut detection uses PySceneDetect; if unavailable, the clip is treated as one shot.
  • The included templates have additional requirements, including VideoHelperSuite for saving.
  • The templates offer automatic ranking or manual face choice per shot.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install the ComfyUI face-refinement nodes

Add the custom nodes to a ComfyUI installation that has MiniMax H3 support.

Before you start

  • ComfyUI with MiniMax H3 support.
  • A face-detection model, such as face_yolov8m.pt, placed in models/ultralytics/bbox/.

Steps

  1. Clone the repository into ComfyUI/custom_nodes/.

  2. Restart ComfyUI.

  3. Find the nodes under MiniMax H3/Face Refine.

Expected outcome

The custom nodes are available in ComfyUI under MiniMax H3/Face Refine.

Guide limitations

  • The face-detection model must be supplied separately.
  • Python dependencies are installed from the project requirements files.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
15 October 2026
Original website title
Carasibana/ComfyUI-H3-FaceRefine · Official homepage title · retained in its source language
How we review sourcesSuggest a correction