Machine Learning has been appearing all over the place these last few years. Google Photos can use it to search for pictures of your cat. Our news apps provide summaries of top stories. Email tools can recognize when we’re writing an upset message and encourage us to think about it before hitting “send.” And life blogging apps track those streams of data our phones create and help us identify that the reason we’re grumpy today is because we haven’t slept well for the last three nights.
These capabilities seem a bit like magic to those of us who are used to traditional programming. And until recently, including these sorts of capabilities in your app required venturing into the deep and scary waters of Tensorflow, Python, regression functions, decision trees, validation sets, and more.
But no more! A few years back, Apple acquired Turi, a machine learning company in Seattle that made high-level ML tools. And before long, Turi’s technology appeared in Apple’s developer tools in the form of a utility called CreateML — an application custom built for generating machine learning models for a variety of tasks: Identifying objects in an image, classifying text, classifying sounds, and identifying activities from motion data. We’ll take a look at the CreateML and its relevant APIs, train an image recognition model, and see how easy it is to add Machine Learning functionality to our apps!
Sean McMains loves to make awesome stuff: cannons, software, trebuchets, games, music, rockets, pranks, hovercraft, and four kids. He wrote his first software for pay at age 12, and has been programming ever since. He’s been fortunate to be a part of the rise of the web, mobile, and now Extended Reality. He’s currently directing the technology implementation team at Handsome in Austin, Texas.