Add AWS Rekognition CustomLabels data loader

[capjamesg]

Description

This PR adds a new data loader that enables loading CustomLabels predictions from AWS Rekognition. This will be useful for people who want to plot bounding boxes or filter predictions with Rekognition CustomLabels.

Note that CustomLabels does not have the concept of a class_id in the output, so my code includes logic to assign those when predictions are ingested into supervision.

Type of change

• New feature (non-breaking change which adds functionality)

How has this change been tested, please provide a testcase or example of how you tested the change?

First, you need to download and install the aws CLI.

How to install the aws CLI.

Then you need to authenticate with the CLI.

Authenticate with the aws CLI.

Then you can run the following code snippet, included in the docstring:

import supervision as sv
import boto3
from PIL import Image

session = boto3.Session()
client = session.client("rekognition")
with Image.open(input) as image:
    buffered = io.BytesIO()
    image.save(buffered, format=image.format)
    image_bytes = buffered.getvalue()

esponse = client.detect_labels(Image={"Bytes": image_bytes})
detections = sv.Detections.from_rekognition_detectlabels(response)

Any specific deployment considerations

N/A

Docs

N/A