Azure Custom Vision Service is a unique venture by Microsoft Azure that seeks to develop computer vision models within a simplistic, drag-and-drop UI. Users can create seamless consumer experiences, improve manufacturing processes, and increase the speed of digital marketing campaigns, among other things. No prior knowledge of machine learning is necessary to leverage this service. Resilient CV models within diverse domains can be built from scratch using the Custom Vision Service.
How to Build a Custom Vision Resource?
In order to create a Custom Vision Project, you first have to create a Custom Vision Resource on your Azure Portal. A Custom Vision resource is created within a Resource Group of your choice. A resource group could be said to be a container to contain multiple Azure Resources.
Note: You must have an Azure Subscription for this. You can opt for the free Azure for Students Version through this link (https://azure.microsoft.com/en-in/free/students/) or access it through the free trial.
Step 1. Log into the Azure Portal with your Azure credentials.
Step 2. Select the “Custom Vision” Service from the list of services displayed.
Step 3. Hit the Create button on the upper ribbon.
Step 4. The “Create Custom Vision” page would be displayed. Fill in the credentials under the Basic tab as:
- Create Options – both (i.e, training and prediction)
- Project details –> Subscriptions –> Navigate to the Subscriptions service to find out your Azure subscription name
- Project details –> Resource Group –> Select a Resource Group of your choice or hit Create new
If you choose the Create new Option, simply enter a unique name for the Resource Group.
- Instance details –> Region –> Closest to you
- Instance details –>Name –> Unique Name for your Instance
- Training Resource –> Training pricing tier –> Standard S0 (10 Transactions per second)
- Prediction Resource –> Prediction pricing tier –> Standard S0 (10k Transactions per month)
Step 5. Click the “Review+Create” button.
Step 6. Once the validation has been passed, Click Create.
Your Custom Vision Resource has been successfully created. You can log onto the official website for custom vision (https://www.customvision.ai/projects) to create customized computer vision models.