The workshop provides 100 pictures of cats and dogs. This is the training data. You use Amazon Rekognition to label them as cat or dog and then train a custom model. You then use the model to identify if any particular picture is of cat or dog programmatically.

Download the training data from the link. Unzip the images and store in some folder on your local computer.

Download the test data from link. Unzip the images and store in some folder on your local computer. You use the test data in your programming to check if the model can detect cat or dog in the picture. You are also free to use your own pictures for the testing.

The workshop got the training and test data from the kaggle.com website.

You will upload the test data images to a S3 bucket.

Login to AWS Console and choose Ireland as the region.

Create a bucket with name dojo-test-images. If this bucket name is not available, use a bucket name which is available. Upload the test images files (test1.jpg,test2.jpg,test3.jpg & test4.jpg) from the test_data.zip file into the bucket created. It has two cat and two dog images for the testing. You can also use your own image for the testing. In that case, upload your image to this S3 bucket.

The test data is ready. The next step is to prepare the training data images.