Building a model to identify Packt’s logo
In this section, we will see how to train a model to identify Packt’s logo using Rekognition Custom Labels. We’ll collect training and test datasets for Packt’s logo, label the images and draw bounding boxes, and train the model.
Step 1 – Collecting your images
Upload the sample images from the book’s GitHub repository. You can complete this step using the following command:
$ aws s3 sync 03_RekognitionCustomLabels/images s3://cv-on-aws-book-xxxx/chapter_03/images --region us-east-2
Important note
To collect the sample images, you can use the same S3 bucket you created in Chapter 2.
Now, navigate to Amazon Rekognition on the AWS Management Console (https://us-east-2.console.aws.amazon.com/rekognition/home?region=us-east-2#/). Select on Use Custom Labels on the left sidebar and then Select on Get started:
Figure 3.1: Amazon Rekognition console
Important...