Amazon Rekognition Custom Labels makes it easy for customers to learn how to train machine learning models by providing tutorial videos, and sample datasets

Amazon Rekognition Custom Labels introduces a simplified on-boarding experience with the ability to explore images, labels, and datasets by one-click creation of example projects. Amazon Rekognition Custom Labels provides out of the box video tutorials, and example projects with hundreds of images for single-class classification, multi-class classification, object detection, and logo detection.

To get started, launch the Rekognition Custom Labels console and select the Get started button. You can select the option to learn through quick two-minute tutorial videos, or select the option to learn through example projects. Once you select an example project, Rekognition Custom Labels automatically creates a new project, relevant test, and training datasets, then creates appropriate labels, and also assigns the labels to the images within the datasets. You can review the images to understand how labels are assigned across classification and object detection use cases. It is also easy to train and evaluate models. All you have to do is click the Train model button on the training dataset page. Typically these models are trained within an hour, and you can get started with evaluating the performance of the newly created Custom Labels model.

Support for this new feature is automatically enabled and available in all Amazon Rekognition Custom Labels regions. To learn more about Amazon Rekognition Custom Labels and region availability, please visit our documentation and region table. 

Get started with Amazon Rekognition Custom Labels today.