Website • Docs • Twitter • Join Slack Community
Label Studio Format Converter helps you to encode labels into the format of your favorite machine learning library.
Running from the command line:
pip install -U label-studio-converter
python label-studio-converter export -i exported_tasks.json -c examples/sentiment_analysis/config.xml -o output_dir -f CSVRunning from python:
from label_studio_converter import Converter
c = Converter('examples/sentiment_analysis/config.xml')
c.convert_to_json('examples/sentiment_analysis/completions/', 'tmp/output.json')Getting output file: tmp/output.json
[
{
"reviewText": "Good case, Excellent value.",
"sentiment": "Positive"
},
{
"reviewText": "What a waste of money and time!",
"sentiment": "Negative"
},
{
"reviewText": "The goose neck needs a little coaxing",
"sentiment": "Neutral"
}
]Use cases: any tasks
Running from the command line:
python label_studio_converter/cli.py --input examples/sentiment_analysis/completions/ --config examples/sentiment_analysis/config.xml --output output_dir --format CSV --csv-separator $'\t'Running from python:
from label_studio_converter import Converter
c = Converter('examples/sentiment_analysis/config.xml')
c.convert_to_csv('examples/sentiment_analysis/completions/', 'output_dir', sep='\t', header=True)Getting output file tmp/output.tsv:
reviewText sentiment
Good case, Excellent value. Positive
What a waste of money and time! Negative
The goose neck needs a little coaxing NeutralUse cases: any tasks
Running from the command line:
python label_studio_converter/cli.py --input examples/named_entity/completions/ --config examples/named_entity/config.xml --output tmp/output.conll --format CONLL2003Running from python:
from label_studio_converter import Converter
c = Converter('examples/named_entity/config.xml')
c.convert_to_conll2003('examples/named_entity/completions/', 'tmp/output.conll')Getting output file tmp/output.conll
-DOCSTART- -X- O
Showers -X- _ O
continued -X- _ O
throughout -X- _ O
the -X- _ O
week -X- _ O
in -X- _ O
the -X- _ O
Bahia -X- _ B-Location
cocoa -X- _ O
zone, -X- _ O
...
Use cases: text tagging
Running from the command line:
python label_studio_converter/cli.py --input examples/image_bbox/completions/ --config examples/image_bbox/config.xml --output tmp/output.json --format COCO --image-dir tmp/imagesRunning from python:
from label_studio_converter import Converter
c = Converter('examples/image_bbox/config.xml')
c.convert_to_coco('examples/image_bbox/completions/', 'tmp/output.conll', output_image_dir='tmp/images')Output images could be found in tmp/images
Getting output file tmp/output.json
{
"images": [
{
"width": 800,
"height": 501,
"id": 0,
"file_name": "tmp/images/62a623a0d3cef27a51d3689865e7b08a"
}
],
"categories": [
{
"id": 0,
"name": "Planet"
},
{
"id": 1,
"name": "Moonwalker"
}
],
"annotations": [
{
"id": 0,
"image_id": 0,
"category_id": 0,
"segmentation": [],
"bbox": [
299,
6,
377,
260
],
"ignore": 0,
"iscrowd": 0,
"area": 98020
},
{
"id": 1,
"image_id": 0,
"category_id": 1,
"segmentation": [],
"bbox": [
288,
300,
132,
90
],
"ignore": 0,
"iscrowd": 0,
"area": 11880
}
],
"info": {
"year": 2019,
"version": "1.0",
"contributor": "Label Studio"
}
}Use cases: image object detection
Running from the command line:
python label_studio_converter/cli.py --input examples/image_bbox/completions/ --config examples/image_bbox/config.xml --output tmp/voc-annotations --format VOC --image-dir tmp/imagesRunning from python:
from label_studio_converter import Converter
c = Converter('examples/image_bbox/config.xml')
c.convert_to_voc('examples/image_bbox/completions/', 'tmp/output.conll', output_image_dir='tmp/images')Output images can be found in tmp/images
Corresponding annotations could be found in tmp/voc-annotations/*.xml:
<?xml version="1.0" encoding="utf-8"?>
<annotation>
<folder>tmp/images</folder>
<filename>62a623a0d3cef27a51d3689865e7b08a</filename>
<source>
<database>MyDatabase</database>
<annotation>COCO2017</annotation>
<image>flickr</image>
<flickrid>NULL</flickrid>
</source>
<owner>
<flickrid>NULL</flickrid>
<name>Label Studio</name>
</owner>
<size>
<width>800</width>
<height>501</height>
<depth>3</depth>
</size>
<segmented>0</segmented>
<object>
<name>Planet</name>
<pose>Unspecified</pose>
<truncated>0</truncated>
<difficult>0</difficult>
<bndbox>
<xmin>299</xmin>
<ymin>6</ymin>
<xmax>676</xmax>
<ymax>266</ymax>
</bndbox>
</object>
<object>
<name>Moonwalker</name>
<pose>Unspecified</pose>
<truncated>0</truncated>
<difficult>0</difficult>
<bndbox>
<xmin>288</xmin>
<ymin>300</ymin>
<xmax>420</xmax>
<ymax>390</ymax>
</bndbox>
</object>
</annotation>Use cases: image object detection
Check the structure of YOLO folder first, keep in mind that the root is /yolo/datasets/one.
/yolo/datasets/one
images
- 1.jpg
- 2.jpg
- ...
labels
- 1.txt
- 2.txt
classes.txt
classes.txt example
Airplane
Car
label-studio-converter import yolo -i /yolo/datasets/one -o ls-tasks.json --image-root-url "/data/local-files/?d=one/images"
Where the URL path from ?d= is relative to the path you set in LABEL_STUDIO_LOCAL_FILES_DOCUMENT_ROOT.
Note for Local Storages
- It's very important to set
LABEL_STUDIO_LOCAL_FILES_DOCUMENT_ROOT=/yolo/datasets(not to/yolo/datasets/one, but/yolo/datasets) for Label Studio to run. - Add a new Local Storage in the project settings and set Absolute local path to
/yolo/datasets/one/images(orc:\yolo\datasets\one\imagesfor Windows).
Note for Cloud Storages
- Use
--image-root-urlto make correct prefixes for task URLs, e.g.--image-root-url s3://my-bucket/yolo/datasets/one. - Add a new Cloud Storage in the project settings with the corresponding bucket and prefix.
Help command
label-studio-converter import yolo -h
usage: label-studio-converter import yolo [-h] -i INPUT [-o OUTPUT]
[--to-name TO_NAME]
[--from-name FROM_NAME]
[--out-type OUT_TYPE]
[--image-root-url IMAGE_ROOT_URL]
[--image-ext IMAGE_EXT]
optional arguments:
-h, --help show this help message and exit
-i INPUT, --input INPUT
directory with YOLO where images, labels, notes.json
are located
-o OUTPUT, --output OUTPUT
output file with Label Studio JSON tasks
--to-name TO_NAME object name from Label Studio labeling config
--from-name FROM_NAME
control tag name from Label Studio labeling config
--out-type OUT_TYPE annotation type - "annotations" or "predictions"
--image-root-url IMAGE_ROOT_URL
root URL path where images will be hosted, e.g.:
http://example.com/images or s3://my-bucket
--image-ext IMAGE_EXT
image extension to search: .jpg, .png
This tutorial will guide you through the process of importing a folder with YOLO annotations into Label Studio for further annotation. We'll cover setting up your environment, converting YOLO annotations to Label Studio's format, and importing them into your project.
- Label Studio installed locally
- YOLO annotated images and corresponding .txt label files in the directory
/yolo/datasets/one. - label-studio-converter installed (available via
pip install label-studio-converter)
Before starting Label Studio, set the following environment variables to enable Local Storage file serving:
Unix systems:
export LABEL_STUDIO_LOCAL_FILES_SERVING_ENABLED=true
export LABEL_STUDIO_LOCAL_FILES_DOCUMENT_ROOT=/yolo/datasets
label-studio
Windows:
set LABEL_STUDIO_LOCAL_FILES_SERVING_ENABLED=true
set LABEL_STUDIO_LOCAL_FILES_DOCUMENT_ROOT=C:\\yolo\\datasets
label-studio
Replace /yolo/datasets with the actual path to your YOLO datasets directory.
- Create a new project.
- Go to the project settings and select Cloud Storage.
- Click Add Source Storage and select Local files from the Storage Type options.
- Set the Absolute local path to
/yolo/datasets/one/imagesorc:\yolo\datasets\one\imageson Windows. - Click
Add storage.
Check more details about Local Storages in the documentation.
Before importing the converted annotations from YOLO, verify that you can access an image from your Local storage via Label Studio. Open a new browser tab and enter the following URL:
http://localhost:8080/data/local-files/?d=one/images/<your_image>.jpg
Replace one/images/<your_image>.jpg with the path to one of your images. The image should display in the new tab of the browser.
If you can't open an image, the Local Storage configuration is incorrect. The most likely reason is that you made a mistake when specifying your Path in Local Storage settings or in LABEL_STUDIO_LOCAL_FILES_DOCUMENT_ROOT.
Note: The URL path from ?d= should be relative to LABEL_STUDIO_LOCAL_FILES_DOCUMENT_ROOT=/yolo/datasets,
it means that the real path will be /yolo/datasets/one/images/<your_image>.jpg and this image should exist on your hard drive.
Use the label-studio-converter to convert your YOLO annotations to a format that Label Studio can understand:
label-studio-converter import yolo -i /yolo/datasets/one -o output.json --image-root-url "/data/local-files/?d=one/images"
Now import the output.json file into Label Studio:
- Go to your Label Studio project.
- From the Data Manager, click Import.
- Select the
output.jsonfile and import it.
After importing, you should see your images with the pre-annotated bounding boxes in Label Studio. Verify that the annotations are correct and make any necessary adjustments.
If you encounter issues with paths or image access, ensure that:
- The LABEL_STUDIO_LOCAL_FILES_DOCUMENT_ROOT is set correctly.
- The
--image-root-urlin the conversion command matches the relative path:
`Absolute local path from Local Storage Settings` - `LABEL_STUDIO_LOCAL_FILES_DOCUMENT_ROOT` = `path for --image_root_url`
e.g.:
/yolo/datasets/one/images - /yolo/datasets/ = one/images
- The Local Storage in Label Studio is set up correctly with the Absolute local path to your images (
/yolo/datasets/one/images) - For more details, refer to the documentation on importing pre-annotated data and setting up Cloud Storages.
We would love to get your help for creating converters to other models. Please feel free to create pull requests.
This software is licensed under the Apache 2.0 LICENSE © Heartex. 2020
![dependabot[bot] avatar](https://avatars.githubusercontent.com/in/29110?v=4 "dependabot[bot]")
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent
