Process a video by using Azure Video Indexer

To process a video using Azure Video Indexer, you can leverage its powerful features to extract insights, generate metadata, and improve video accessibility. In this article, we will explore the steps involved in designing and implementing a Microsoft Azure AI solution using Azure Video Indexer.

Step 1: Set up Azure Video Indexer

1. Create an Azure Video Indexer resource in the Azure portal. Select the appropriate subscription, resource group, and region.
2. Configure the details of the Video Indexer resource, such as the name, pricing tier, and storage account. You can choose between a free or standard tier, depending on your requirements and expected usage.
3. Wait for the deployment to complete, and then navigate to the newly created Video Indexer resource in the Azure portal.
4. Obtain the necessary API keys for accessing the Video Indexer service. Go to the “Keys and Endpoint” section in the Video Indexer resource and note down the API subscription key and the endpoint URL. These will be used in the following steps.

Step 2: Upload and Process a Video

Now that you have set up Azure Video Indexer, let’s proceed with uploading and processing a video:

1. Gather the video file that you want to process. Ensure it is in a supported format (e.g., MP4).
2. Use the Video Indexer API to programmatically upload the video for processing. You can make a POST request to the https://api.videoindexer.ai/{location}/Accounts/{accountId}/Videos?accessToken={accessToken}&name={videoName}&privacy={privacy} endpoint, replacing the placeholders with appropriate values.
3. You can obtain the access token by making a request to the https://api.videoindexer.ai/auth/{location}/Accounts/{accountId}/AccessToken?allowEdit=true endpoint, passing the location and accountId in the URL.
4. After successfully uploading the video, you will receive a response with a video id. Make note of this id as it will be used to fetch the processed video insights.

Step 3: Retrieve Video Insights

Once the video has been uploaded and processed, it’s time to retrieve the insights generated by Azure Video Indexer:

1. Utilize the Video Indexer API to fetch the insights for a specific video. Make a GET request to the https://api.videoindexer.ai/{location}/Accounts/{accountId}/Videos/{videoId}/Index?accessToken={accessToken} endpoint, replacing the placeholders with the appropriate values.
2. Parse the response to extract relevant information such as transcript, keywords, sentiment analysis, named entities, and scene descriptions. These insights are generated by Azure Video Indexer’s AI models and can help you gain a deeper understanding of the video content.
3. Leverage the extracted insights to enhance video accessibility, enable content discovery, or perform further analysis.

Step 4: Customize AI Models (Optional)

Azure Video Indexer provides the ability to customize AI models for specific scenarios. This step is completely optional but can be valuable if you have unique requirements. Follow these steps to customize the AI models:

1. Navigate to the Azure Video Indexer portal and sign in using your Azure credentials.
2. Select the desired Video Indexer account and click on the “Customization” tab.
3. Create a new customization project and follow the instructions provided to upload training data, train the model, and evaluate its accuracy.
4. Once the model is trained, you can apply it to your videos to obtain customized insights.

That’s it! You have now learned the process of using Azure Video Indexer to process a video, extract insights, and improve video accessibility using AI. Azure Video Indexer offers powerful capabilities to enhance your videos and unlock valuable information.

Feel free to explore the Azure Video Indexer documentation for detailed information on the API endpoints, supported features, and customization options.