Quick Start Guide¶

Get up and running with Labelformat in minutes! This Quick Start Guide provides simple, copy-paste examples to help you convert label formats effortlessly.

Scenario 1: Convert COCO to YOLOv8 Using CLI¶

Step 1: Prepare Your Files¶

Ensure you have the following structure:

project/
├── coco-labels/
│   └── train.json
├── images/
│   ├── image1.jpg
│   └── image2.jpg

Step 2: Run the Conversion Command¶

Open your terminal, navigate to your project directory, and execute:

labelformat convert \
    --task object-detection \
    --input-format coco \
    --input-file coco-labels/train.json \
    --output-format yolov8 \
    --output-file yolo-labels/data.yaml \
    --output-split train

Step 3: Verify the Output¶

Your project structure should now include:

project/
├── yolo-labels/
│   ├── data.yaml
│   └── labels/
│       ├── image1.txt
│       └── image2.txt

Scenario 2: Convert YOLOv8 to COCO Using Python API¶

Step 1: Install Labelformat¶

If you haven't installed Labelformat yet, do so via pip:

pip install labelformat

Step 2: Write the Conversion Script¶

Create a Python script, convert_yolo_to_coco.py, with the following content:

from pathlib import Path
from labelformat.formats import COCOObjectDetectionOutput, YOLOv8ObjectDetectionInput

# Load YOLOv8 labels
yolo_input = YOLOv8ObjectDetectionInput(
    input_file=Path("yolo-labels/data.yaml"),
    input_split="train"
)

# Convert to COCO format and save
coco_output = COCOObjectDetectionOutput(
    output_file=Path("coco-from-yolo/converted_coco.json")
)
coco_output.save(label_input=yolo_input)

print("Conversion from YOLOv8 to COCO completed successfully!")

Step 3: Execute the Script¶

Run the script:

python convert_yolo_to_coco.py

Step 4: Check the COCO Output¶

Your project should now have:

project/
├── coco-from-yolo/
│   └── converted_coco.json

Scenario 3: Convert Labelbox Export to Lightly Format¶

Step 1: Export Labels from Labelbox¶

Ensure you have the Labelbox export file, e.g., labelbox-export.ndjson.

Step 2: Run the Conversion Command¶

labelformat convert \
    --task object-detection \
    --input-format labelbox \
    --input-file labelbox-export.ndjson \
    --category-names cat,dog,fish \
    --output-format lightly \
    --output-folder lightly-labels/annotation-task

Step 3: Verify the Lightly Output¶

Your project structure should include:

project/
├── lightly-labels/
│   ├── annotation-task/
│   │   ├── schema.json
│   │   ├── image1.json
│   │   └── image2.json