In the previous section, we deployed a website to Azure that displays all the images in our blog storage. The goal of this step is to create a Cognitive Service, and use its Vision API to detect dogs in our images.
The Cognitive Services are a great way to infuse your apps, websites and bots with intelligent algorithms to see, hear, speak, understand and interpret your user needs through natural methods of communication. Transform your business with AI today.Learn more...
The most common tool when we start with Azure, or when we explore new features is the Azure portal. It's always updated and very easy to use.
Navigate to https://portal.azure.com
Let's do a quick tour.
In the top left corner (A) you have:
- The big green "+" button to create any resources.
- The home button that brings you back to this view.
- The dashboard button that brings you to your custom dashboards that you can share with others.
- All Services will display the list of all the services grouped by families. From there you can narrow the selection and see a list of your resources matching your type selection.
Just under it, you have the favorites (B). This list is completely configurable. Just like bookmarks, by clicking on one of them you will see all the instances of a certain type listed for you.
In the top of the screen in the middle there is the search bar (C). It will search through types, instances and even documentation!
Just on the right side of the search bar you have a few useful buttons:
- Cloud Shell, a fully functional terminal
- A filter to change what you see in the portal. Very useful when you have access to multiple subscriptions.
- The notification list.
- Settings, where you can change the language, color, of the portal, besides many other things
- Help, and Feedback
Finally, in the top right corner it's you (E) and a way to access all the account settings.
Now that we know the portal, let's create a Cognitive Service instance.
-
Click the "+" button from the top left corner.
-
In the search bar, type vision
-
From the list select Computer Vision
-
Click the Create button.
It's now time to select the properties of your Vision API.
- Give it a unique name. Be sure there is a little green check at the right of the textbox, meaning that it is valid.
- Validate if it's in the correct subscription. It should be the same used previously to deploy the website.
- You can select any location. It's recommended to select the same for all the resources to get the best performance.
- Select F0 (aka Free) for the Pricing tier.
- Select the same resource group that was created for the previous deployment.
It should look like this. Click the Create button.
After a few seconds, the Cognitive Services - Computer Vision resource should be created.
To be able to use this API we will need to get its endpoint and the key.
-
From the left panel, click on Resource groups. This will list all the groups in your subscription.
-
Click on your resource group (ex: Hackdemo).
-
Once more, we are in the resource group with the website, storage account and service plan. This time there is a new resource available. Click on your Computer Vision (ex: smartVision)
-
From the left panel select Overview. In the top-right section of that blade, you will find the Endpoint. Note that, we will need it later.
-
Now we need the access Key. To find it, select Key from the left panel, and note one of the two keys.
You have two keys for security purposes. This way you can update your security without stopping your services.
You have now completed this part of the workshop. You can continue with Part 3: Build and Deploy The Azure Function.
In this workshop we used one of the Vision APIs available in the Azure Cognitive Services. However, the Azure Cognitive Services offer many other services. Some services are even available in Docker Containers!
-
Decision
Build apps that surface recommendations for informed and efficient decision-making.
-
Speech
Convert speech into text and text into natural-sounding speech. Translate from one language to another and enable speaker verification and recognition.
-
Language
Allow your apps to process natural language with pre-built scripts, evaluate sentiment and learn how to recognize what users want.
-
Search
Add Bing Search APIs to your apps and harness the ability to comb billions of webpages, images, videos, and news with a single API call.
-
Vision
Recognize, identify, caption, index, and moderate your pictures, videos, and digital ink content.
You can get started for free, see interactive demos and even get some starter kits here