Tutorial

Setting the credentials

import spb_curate

spb_curate.access_key = "..."
spb_curate.team_name = "..."

Create your first dataset

By the end of this tutorial you will have created a dataset and added an image (via local storage or URL) with an annotation.

import time

from spb_curate import curate

dataset = curate.create_dataset(
    name="My first dataset",
    description="For setting up the Superb AI Curate Python client"
)

image_1_key = "unique-image-key-1"
image_2_key = "unique-image-key-2"

images = [
    # Add image from local storage
    curate.Image(
        key=image_1_key,
        source=curate.ImageSourceLocal(asset="/path/to/image"),
        metadata={"weather": "clear", "timeofday": "daytime"},
    ),

    # Add image from URL
    curate.Image(
        key=image_2_key,
        source=curate.ImageSourceUrl(url="http://example.com/path/to/image.jpg"),
        metadata={"weather": "sunny", "timeofday": "daytime"},
    ),
]

job = dataset.add_images(images=images)
job.wait_until_complete()

time.sleep(5)

box_x, box_y, box_w, box_h = (0, 0, 100, 100)

annotations = [
    curate.Annotation(
        image_key=image_1_key,
        annotation_class="dog",
        annotation_value=curate.BoundingBox(
            x=box_x, y=box_y, width=box_w, height=box_h),
        metadata={"iscrowd": "true"},
    )
]

job = dataset.add_annotations(annotations=annotations)
job.wait_until_complete()