forked from carbaro/label-studio-converter
-
Notifications
You must be signed in to change notification settings - Fork 0
Home
Will O'Brien edited this page Feb 16, 2024
·
8 revisions
Welcome to the label-studio-converter wiki! This wiki page highlights the specific uses of this repository for AppMAIS.
This repository is used for importing pre-labeled images in the YOLO format into label studio. This is done in a few steps:
- Using serve-data.py to host the image data locally.
- Converting yolo labels to label studio format for Segmentation models and for Detection models.
- Cleaning the data and then importing the images and labels into label studio
If you haven't accessed the label studio front end before, we host it on the lambda machine on port 8000. To access it, you have to use the -L
flag with ssh
.
ssh -L 8000:localhost:8000 <user>@<ip>
Which can be used multiple times if multiple ports are needed, such as if you needed port 3030 as well as 8000:
ssh -L 8000:localhost:8000 -L 3030:localhost:3030 <user>@<ip>
Currently, port 8000 is used for the label studio front end and port 3030 is used for local image hosting.