Start using Oracle Analytics and AI in 5 minutes!

September 20, 2022 | 4 minute read
Benjamin Arnulf
Senior Director, Product Strategy, Oracle Analytics
Text Size 100%:
AI Vision and Oracle Analytics
AI Vision using Oracle Analytics

Oracle Analytics integrated with AI Vision Services with object detection.

Oracle Analytics is now integrated with Oracle Artificial Intelligence Vision service. I recently created a personal free Oracle Cloud account to try the new Artificial Intelligence Vision service. The goal was to test the integration with Oracle Analytics. Here is how you can do the same and create an Analytics and AI data visualization in 5 minutes!

  1. Create your Oracle Cloud account.

Use your current OCI account or create one. You can create a free tier Oracle Cloud account. Follow this link to get started: https://www.oracle.com/cloud/free/

 

  1. Create your Oracle Analytics instance.

AI Analytics Services

Oracle Cloud screen to create an Oracle Analytics instance, Object Store, and more.

To create a data visualization, use an Oracle Analytics account. You can use your current account or create one following the steps below:

Go to Oracle Cloud: https://www.oracle.com/cloud/sign-in.html

Click the Analytics & AI menu and create your instance.

Sign in to Oracle Analytics and add your Oracle Cloud OCI domain to the list of safe domains.

Navigate to the console and click Safe Domains.

 

  1. Create your Object Storage Bucket.

In Oracle Cloud, click the Storage menu and select Buckets.

Create the bucket to store your images.

Ensure your bucket storage is set to Public.

 

  1. Upload your images.

From OCI, upload your images to the bucket with their original extensions.

Save the URL Path of your object storage (Using the View Object Details menu).

Copy the images list from the bucket and create a .CSV file with the following format:

Column A: Id (example: 1, 2, 3)

Column B: Image Name (example: Boat.jpg)

Column C: File Location (example: https://objectstorage.us.....)

Column D: Uploaded Date (example: Fri, Jul 1, 2022, 04:00:00 UTC)

Column E: Size (example: 500 KIB)

Column F: Type (example: Standard)

 

  1. Create an OAC connection.

Log in to OCI and copy your Tenancy OCID and User OCID.

In Oracle Analytics, create an OCI Resource connection.

Use your Tenancy OCID and User OCID.

Click Generate Key and add the key to API keys in OCI in User Settings.

 

  1. Register your AI Vision model.

In Oracle Analytics, click the 3 dots close to your profile picture.

Select Register Model/Function and OCI Vision Models.

Select the OCI connection you created, then select Pretrained Object Detection.

Add the name of your bucket to register it.

 

  1. Apply your AI vision model.

 

Apply AI Vision Model in Data Flow for Analytics

Picture of Oracle Analytics applying an AI model in Data Flow.

In Oracle Analytics, create a data flow.

Add data by selecting your .CSV file.

Apply the AI model.

The input column should be the file location.

The input type should be images.

Add a Save Data step and save it as a dataset.

Run your AI Vision model.

The AI model detects objects and stores the detected objects in the dataset.

 

  1. Create an Analytics data visualization using AI.

Download the AI vision plugin for Oracle Analytics.

In Oracle Analytics, add your new dataset that was generated with the data flow.

Create a table with object name, vertex 1, vertex 2, image name.

Make this table a filter by clicking the top left filter icon on the visualization.

Drag and drop the Vision Plugin data visualization in your workbook.

Populate the image location and include 2 vertex columns. Then click the image in the table.

You should now see your images on the Vision Plugin visualization.

AI Vision RedBull
AI Vision RedBull - Author: Benjamin Arnulf

If you have issues seeing your images and the object detected:

In Google Chrome, open the browser and enter this URL:
chrome://flags/#block-insecure-private-network-requests

You might need the following extension with Google Chrome:

https://chrome.google.com/webstore/detail/always-disable-content-se/ffelghdomoehpceihalcnbmnodohkibj?utm_source=chrome-ntp-icon

Also temporarily disable “Block insecure private network requests”, if you have issues accessing your object storage as a debugging test.

 

For more information about Oracle Analytics go to https://www.oracle.com/business-analytics/

For more information about Oracle AI go to https://www.oracle.com/artificial-intelligence/

Benjamin Arnulf

Senior Director, Product Strategy, Oracle Analytics

Benjamin Arnulf is Senior Director, Product Strategy covering Oracle Analytics and AI.


Previous Post

Enabling Extensive Flexfields in Fusion Analytics Warehouse Part 2

Krishna Prasad Kotti | 4 min read

Next Post


Make Rich Data Even Richer – Use Functions for Custom Data Enrichment in Oracle Analytics Cloud

Katherine Deegan | 3 min read