From photos of earth from space to cellular microscopy and everything in between, we’ve seen our fair of images come through CrowdFlower over the years. Millions of them, in fact. And we’ve learned a few things along the way.
We’ve learned how best to scope an image annotation project. We’ve learned which tools work best for which job. We’ve learned how smart projects succeed and how other ones fall down because of small oversights. We’ve learned how to solve for overfitting, why simple classification is sometimes just as successful as pixel-by-pixel labeling, and how much training data you need to make your project a success.
We’ve compiled that information in a free eBook, which you can download below. And if you’re thinking about launching a computer vision project–or if you just want to learn a little bit about how they work–we think you’ll learn a ton of relevant if you do.