GettingStarted.md

Gulla.Episerver.AutomaticImageDescription

1. Install the addon

The NuGet package is available at https://nuget.optimizely.com/

2. Configuration

For AutomaticImageDescription to work, you will have to call the .AddAutomaticImageDescription() extension method in the Startup.ConfigureServices method. You may pass configuration to the .AddAutomaticImageDescription() method or use appsettings.json instead. A configuration setting specified in appsettings.json will override any configuration configured in Startup.cs.

3. Create a Computer Vision resource

In the Azure portal, create a new resource. Search for «Computer Vision», select pricing tier and create.

4. Add Computer Vision Key and Endpoint to config

After the resource is created, locate «Keys and Endpoint» in the left pane and add them to appsettings.json like this.

"Gulla": {
  "AutomaticImageDescription": {
    "ComputerVisionSubscriptionKey": "key1",
    "ComputerVisionEndpoint": "https://myendpoint.cognitiveservices.azure.com/"
  }
}

5. Create a Translator resource

If you are happy with English metadata, you may skip step 3 and 4. If you want your metadata translated to other languages you will need to create a Translator resource. Search for «Translator», select pricing tier and create.

6. Add Translator Key to config

After the resource is created, locate «Keys and Endpoint» in the left pane and add the key to the appsettings.json like this.

"Gulla": {
  "AutomaticImageDescription": {
    "TranslatorSubscriptionKey": "key2"
  }
}

7. Add Translator Region to config (optional)

Add your location, also known as region. The default is global. This is required if using a Cognitive Services resource.

"Gulla": {
  "AutomaticImageDescription": {
    "TranslatorSubscriptionRegion": "YOUR_RESOURCE_LOCATION"
  }
}

Summary, all possible configuration

If you prefer appsettings.json:

"Gulla": {
  "AutomaticImageDescription": {
    "ComputerVisionSubscriptionKey": "key1",
    "ComputerVisionEndpoint": "https://myendpoint.cognitiveservices.azure.com/"
    "TranslatorSubscriptionKey": "key2"
    "TranslatorSubscriptionRegion": "YOUR_RESOURCE_LOCATION",
    "ScheduledJobMaxRequestsPerMinute": 10
  }
}

If you prefer startups.cs:

.AddAutomaticImageDescription(x =>
{
    x.ComputerVisionSubscriptionKey = "key1";
    x.ComputerVisionEndpoint = "https://myendpoint.cognitiveservices.azure.com/";
    x.TranslatorSubscriptionKey = "key2";
    x.TranslatorSubscriptionRegion = "YOUR_RESOURCE_LOCATION";
    x.ScheduledJobMaxRequestsPerMinute = 10;
})

Now, you are ready to generate metadata! The next step is to add some Attributes to your image model!

<< Back to readme