Dan04ggg

Dan04ggg / VisOS

Public

Tired of paying for expensive computer vision dataset management platforms? VisOS offers a free local, dataset annotation, management and no code training platform for computer vision models

41
6
100% credibility
Found Apr 16, 2026 at 41 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

VisOS is a local web app for managing computer vision datasets with tools for annotation, format conversion, augmentation, merging, duplicate removal, video extraction, and model training.

How It Works

1
🔍 Discover VisOS

You find VisOS, a friendly tool that handles all your image labeling chores without needing the cloud or any coding.

2
🚀 Start with one command

Run a simple starter script and everything launches automatically in your browser, ready to use in minutes.

3
📁 Load your images

Drop in your photos or videos, and VisOS instantly recognizes the format and shows you what's inside.

4
✨ Annotate effortlessly

Draw boxes, shapes, or use smart auto-tools to label objects quickly, with undo and easy class management.

5
🔄 Prepare and expand

Convert formats, remove duplicates, augment data, or split into training sets with simple clicks.

6
⚡ Train your model

Pick a ready model, hit train, and watch live charts as it learns from your labeled images.

🎉 Custom model ready

Your personalized AI vision model is trained locally and ready to detect objects just like you taught it.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 41 to 41 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 VisOS?

VisOS is a free local web app for managing computer vision datasets—load folders or ZIPs in 15+ formats like YOLO, COCO, VOC, auto-detect structure, then annotate with bbox/poly/keypoint/SAM tools, convert formats, augment with 25+ transforms, merge datasets, extract video frames, detect duplicates, and train models like YOLOv8-12 or SAM directly. Tired of paying bills for cloud platforms? Fire up `python run.py restart` for a Next.js frontend on :3000 proxying a FastAPI Python backend on :8000, with PyTorch handling inference and training—no accounts, uploads, or code needed.

Why is it gaining traction?

It bundles every CV dataset chore into one offline UI: class merging/renaming without JSON hacks, live training metrics with pause/resume/export to ONNX, keyboard-driven sorting/filtering, and batch auto-annotation jobs. Tired of paying for binge full episodes of format converters on GitHub? Developers grab it for the no-code polish—previews before augmenting, stratified splits, YAML editors—that saves hours vs. scripting.

Who should use this?

CV engineers tired of paying rent on expensive annotation platforms, ML researchers prototyping YOLO/SAM models locally without AWS bills, teams merging messy multi-format datasets from Kaggle/competitions. Skip if you need team collab or massive scale.

Verdict

Promising local CV workbench with stellar docs and one-command setup, but 41 stars and 1.0% credibility signal early maturity—test on small datasets first. Worth forking if you're tired of paying taxes for cloud tools; production? Wait for more adoption.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.