Before you begin
Set up your Google Cloud Platform project, authentication, and enable AutoML Vision.
Preparing your training data
Learn best practices in organizing and annotating the images you will use to train your model, as well as format a training CSV file.
Creating datasets and importing images
Create the dataset and import the training data used to train your model.
Training Cloud-hosted models
Train your custom model hosted on the Cloud and get the status of the training operation.
Training Edge (exportable) models
Train your custom exportable Edge model and get the status of the training operation.
Review the performance of your model.
Deploy your model for use after training completes.
Making individual predictions
Use your custom model to annotate an individual prediction image with labels and bounding boxes online.
Making batch predictions
Use your custom model to annotate a batch of prediction images with labels and bounding boxes online.
Exporting Edge models
Export your different trained Edge model formats to Google Cloud Storage and for use on edge devices.
Undeploy your model after you are done using them to avoid further hosting charges.
Manage datasets associated with your project.
Manage your custom models.
Working with long-running operations
Get the status of long-running operations.
Use native "base64" utilities to encode a binary image into ASCII text data to send in an API request.
Filtering when listing
Learn how to filter results when listing resources, operations, and metrics.
Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License. For details, see the Google Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates.