ipyezannotation 0.2.5

Easy, simple to customize, pythonic data annotation framework.

Easy Annotation

ipyezannotation - Easy, simple to customize, pythonic data annotation framework.

Disclaimer

This project is in early development stage, so don't blame me if it opens-up a black hole in your HDD 😄, other than that IT WORKS! 🥳

Docs & examples coming soon.

Dependencies

This project currently supports python>=3.8. In future version of this project (possibly ipyezannotation>=1.0.0) only later python versions will be supported starting from 3.9 or 3.10.

Installation

There are two options to install this project:

• Download and install from PyPI by simply running: pip install ipyezannotation & you're done!
• Alternatively, install from source using Poetry. This project uses poetry>=1.3 to manage dependencies.

Examples

Images selection annotation

Annotation using ImageSelectAnnotator.

Define data to annotate with ImageSelectAnnotator:

source_groups = [
    ["./surprized-pikachu.png"] * 16,
    ["./surprized-pikachu.png"] * 7,
    ["./surprized-pikachu.png"] * 8,
    ["./surprized-pikachu.png"] * 4,
]

Convert input data to Sample's:

from ipyezannotation.studio import Sample, SampleStatus

samples = [
    Sample(
        status=SampleStatus.PENDING,
        data=image_paths,
        annotation=None
    )
    for image_paths in source_groups
]

Initialize database of your liking and synchronize it with your new input samples:

from ipyezannotation.studio.storage.sqlite import SQLiteDatabase

db = SQLiteDatabase("sqlite:///:memory:")
synced_samples = db.sync(samples)

Configure & create annotation Studio to label your samples:

from ipyezannotation.studio import Studio
from ipyezannotation.annotators import ImageSelectAnnotator

Studio(
    annotator=ImageSelectAnnotator(n_columns=8),
    database=db
)

Inspiration

Love letter to the following projects coming soon ❤️

• ipyannotations
• superintendent
• label-studio