Label Studio Introduction

Preparing a Dataset with Label Studio

This guide explains how to create an annotated dataset for YOLO training using Label Studio.

Install Label Studio

Refer to the official installation guide: https://labelstud.io/guide/install

Enable Local File Storage

For large projects it is not recommended to upload images directly through the Label Studio interface.

To enable local file access, configure the following environment variables:

LABEL_STUDIO_LOCAL_FILES_SERVING_ENABLED=true
LABEL_STUDIO_LOCAL_FILES_DOCUMENT_ROOT=/path/to/images

Launch Label Studio:

label-studio start

Create a Project

1. Login to Label Studio
2. Click Create Project
3. Enter a project name
4. Configure the labeling interface
5. Save the project

Import Images

1. Open the project
2. Click Import
3. Select Local Storage

When importing images, choose Files as the import method.

Unlike CVAT, Label Studio creates one task for each imported document image.

File:LabelStudio-import-method.png

Selecting the Files import method

Annotate Documents

1. Open a task
2. Select a label
3. Draw a bounding box around the target area
4. Save the annotation

Example labels:

• Invoice Number
• Date
• Seller Name
• Buyer Name
• Total Amount

File:LabelStudio-annotation-example.png

Example annotation

Export the Dataset

1. Open the project
2. Click Export
3. Select the desired format

Supported formats include:

• YOLO
• COCO
• Pascal VOC
• CSV

For YOLO training, export the dataset in YOLO format.

Dataset Formats

COCO

COCO is a JSON-based dataset format commonly used for object detection datasets.

More information: https://docs.aws.amazon.com/rekognition/latest/customlabels-dg/md-coco-overview.html

YOLO

YOLO datasets contain images and annotation files organized according to a predefined directory structure.

More information: https://docs.cvat.ai/docs/dataset_management/formats/format-yolo/

YOLOv8 OBB

YOLOv8 OBB (Oriented Bounding Boxes) extends the standard YOLO format by supporting rotated bounding boxes using eight normalized coordinates.