In this part of the challenge we will choose the images we want Amazon Rekognition to use to train our machine learning model. We need to first create a dataset. The dataset will be a collection of puppy photos that Amazon Rekognition will learn from in order to detect puppies in photos.
To train Amazon Rekognition how to detect puppies, we need to give it a set of puppy photos first so it can learn what they look like.
You have ten minutes to find the following photos of puppies on the internet. Search the free stock photos Pexels website to find some photos of puppies.
Images must be less than 4096 pixels and greater than 64 pixels. Many of the Pexels photos are larger than this and you might need to resize your images.
You can also download these resized images. Download a collection of photos of puppies from Pexel
Make sure to save all pictures to your desktop.
Like when you have a puppy, it is also important to name our data set. Type in PuppyPhotos into the data set name box.
We need to upload the pictures you found earlier to the data set! Choose Upload images from your computer.
A Tool guide popup may appear the first time you create a dataset. Choose Next until the popup closes.
Choose + Add images
To add photos into the dataset, you can either drag your photos from the desktop into the data set or click Choose files and select your puppy photos from the desktop.
You can add 30 images this way at a time.
You might get an error because your photos are too large or a problem with the filename. If you have problems, try uploading one image at a time. Images must be less than 4096 pixels and greater than 64 pixels. Try a smaller image if you have problems uploading larger images. Or Download a collection of photos of puppies from Pexel
Choose Upload images.
In the next step, we will start labelling our photos.