Oracle Cloud Infrastructure Documentation
Try Free Tier

Building a Custom Model in the Console

Vision provides an option to build custom models to extract insights from images, without needing Data Scientists.

In this tutorial you learn how to:

Before You Begin

Before you use Vision, your tenancy administrator must set up the appropriate policies.

Setting Up the Policies

Follow these steps to set up the policies needed to use Vision.

  1. In the Console navigation menu, select Identity & Security.
  2. Under Identity, select Policies.
  3. In the Policies page, select Create Policy.
    The Create Policy panel is displayed.
  4. Enter a Name. You can use alphanumeric characters, hyphens, periods, and underscores only. Spaces aren't allowed. For example, enter vision-access-policy.
  5. Enter a Description to help other users know the purpose of this set of policies. For example, enter Policy to access Vision service..
  6. Select the Compartment.
  7. In Policy Builder, select Show manual editor.
  8. Add the following statement: 
    allow any-user to use ai-service-vision-family in tenancy
  9. (Optional) To limit access to your user group only, add the following policy instead: 
    allow group <your-group-name> to use ai-service-vision-family in tenancy
  10. Select Create.

1. Create a Project

A Project is a way to organize many models in the same workspace.

Creating a Project

Follow these steps to create a Project in Vision.

  1. In the Console navigation menu, select Analytics & AI.
  2. Under AI Services, select Vision.
  3. From the Vision Console page, under Custom Models, select Projects.
    The Project List page is displayed.
  4. Select Create Project.
    The Create Project panel is displayed.
  5. Select a Compartment to create the Project in.
  6. Give the Project a Name. For example vision_demo. Don't enter confidential information.
  7. (Optional) Enter a Description for the Project to help others identify it.
  8. Select Create Project.

2. Create and Train a Custom Model

Follow these steps to create a custom model in your project.

Complete 1. Create a Project before trying this section.

Select the Model Type

  1. Under Choose Model Type to Train, select Type.
  2. Select a model type from the list.
    Figure 1. Select Model Type
    Select Model Type to Train list with Image Classification selected.

Select the Training Dataset

  1. If you have no annotated images, click Create a New Dataset. You're taken to Oracle Cloud Infrastructure Data Labeling to create new dataset. More information on annotating images is available in the Data Labeling documentation.
  2. If you have a dataset with annotated images, click Choose Existing Dataset.
    1. If you annotated the images in Data Labeling, click Data Labeling Service and select your dataset file.
    2. If you annotated the images in a third-party tool, click Object Storage and select your dataset file.
    Figure 2. Training Data Selections
    Select Existing Dataset selected, as is Data Labeling Service. A dataset is selected from the list.

Train the Custom Model

  1. Enter a Name for the model.
  2. Enter a Description so that you and other users can easily find the model.
  3. Specify a training duration.
    Figure 3. Training Durations
    The model training duration options, Recommended training, Quick train, and Custom. Recommended training is selected.
  4. Select Next.
  5. Review the model information.
  6. Select Submit to start training the model.

What's Next

Now you know how to use Vision with custom models, try using it with pretrained models.

Was this article helpful?