Aero-Ex

Aero-Ex / ComfyUI-SegviGen

Public

ComfyUI-SegviGen: A ComfyUI implementation of SegviGen, providing precise 3D part segmentation for SegviGen.

18
4
100% credibility
Found Mar 25, 2026 at 18 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

ComfyUI custom nodes that use AI to segment 3D models into distinct parts guided by a reference image.

How It Works

1
🔍 Discover SegviGen in ComfyUI

While building AI art workflows in ComfyUI, you find these new nodes that promise to magically split 3D models into parts using just a photo.

2
📥 One-click setup

Drag the folder into your ComfyUI custom nodes, run the easy install script, and watch models download automatically.

3
🧩 Load your 3D model and photo

Pick your GLB file and a reference image showing the colors or style you want for each part.

4
✨ AI segments it perfectly

Connect a few nodes, hit queue prompt, and see your model break into beautifully colored, distinct parts matching the photo.

5
Pick your perfect output
🖼️
Bake onto original mesh

Keep your exact shape with fresh textures applied seamlessly.

🔄
Generate shiny new version

Create a high-quality remeshed model with perfect details.

6
✂️ Split into separate pieces

Use the splitter node to export individual parts as their own files if you want to edit them alone.

✅ Your segmented 3D model shines!

Download the colorful, ready-to-use GLB and drop it into your 3D software, game engine, or viewer.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 18 to 18 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is ComfyUI-SegviGen?

ComfyUI-SegviGen is a Python implementation providing precise 3D part segmentation for SegviGen right inside ComfyUI workflows. Upload a textured GLB mesh and a reference image, and it preprocesses the image (background removal via BiRefNet), conditions with DinoV3, samples textures via flow-matching, then outputs retextured voxel or GLB parts matching the image's style. Users get modular nodes for loaders, samplers, and exporters, with auto model downloads from Hugging Face.

Why is it gaining traction?

This ComfyUI-SegviGen setup shines with VRAM-smart decoupled encoders/decoders, DMA loading, and cache clearing, handling TRELLIS 2.0's 4B params without OOM on mid-tier GPUs. Granular nodes let you chain preprocessing to GLB baking or splitting, skipping manual scripting for pro segmentation results. The install.py script grabs custom CUDA wheels, making setup painless despite dependencies.

Who should use this?

ComfyUI users in 3D asset pipelines needing image-guided part segmentation, like character retexturing or product mockups from reference photos. 3D ML devs prototyping SegviGen variants in visual graphs. Skip if you're CLI-only or outside ComfyUI.

Verdict

Grab it for ComfyUI-SegviGen if precise part segmentation slots into your nodes—works as advertised post-install. At 18 stars and 1.0% credibility, it's early (README docs only, no tests), so test on non-critical workflows and watch for updates.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.