The easiest & fastest way to annotate images for vision AI. | Created by @wkentaro_ | 16K GitHub stars
Joined November 2023
- Tweets399
- Following2
- Followers340
- Likes1.3K
Pinned Tweet
Labelme is now on the Microsoft Store.
Click Get, then Open, and you can start annotating. Windows doesn't show the blue warning screen anymore.
apps.microsoft.com/detail/9M…
Labelme retweeted
I can finally sleep without worrying that SmartScreen is stopping people from opening Labelme.
Labelme is now on the Microsoft Store.
Click Get, then Open, and you can start annotating. Windows doesn't show the blue warning screen anymore.
apps.microsoft.com/detail/9M…
If your IT team installs software for you, they can use winget:
winget install --id 9MSSLKN1F14Q --source msstore
On macOS or Linux, download it from labelme.io
On labelme.io, the Windows download now goes to the Microsoft Store. The .msi installer is still there if you need it.
ALT labelme.io download page. Under Windows, the Microsoft Store button is circled, with Installer (.msi) below it.
Labelme retweeted
We paid for a code-signing certificate and signed the Labelme installer. Windows still showed new users a blue "Windows protected your PC" screen, and some gave up there.
The Microsoft Store version installs without it.
Labelme is now on the Microsoft Store.
Click Get, then Open, and you can start annotating. Windows doesn't show the blue warning screen anymore.
apps.microsoft.com/detail/9M…
Labelme retweeted
SAM3 now runs offline on your GPU.
An ONNX export of Meta's SAM3 that runs on NVIDIA CUDA and Apple CoreML. Besides text, it takes several example boxes, negative boxes and clicks as prompts.
A text prompt takes 0.37 s on an M5 MacBook Pro and 0.08 s on an RTX 3080 Ti.
We'll soon bring this into Labelme so you get a much smoother local annotation experience
ALT Bar chart of one SAM3 text prompt: 1.43 s on Apple M5 CPU, 373 ms on M5 CoreML, 77 ms on RTX 3080 Ti CUDA.
Labelme retweeted
annotation is turning into coding:
1. nail the spec on a few images
2. batch annotate
3. review, by you or another model
4. merge into your dataset
Labelme retweeted
same model, same effort. only the prompt changed.
the right prompt tells it to zoom in, leave out what covers the backpack, and use group_id for shapes with holes.
Blog: labelme.io/blog/ai-agents
Setup guide: labelme.io/docs/connect-ai-a…
Labelme retweeted
Shape locking and Fit to Selection are now in Labelme. Both great for reviewing what you or your agents or AI models annotated.
pip install -U labelme
Lock Shapes: labelme.io/blog/lock-shapes
Fit to Selection: labelme.io/blog/fit-to-selec…
pip install -U labelme
