Azure AI Services at Scale for Cloud, Mobile, and Edge

Chapter 11. Translating Multiple Languages at Scale for International Organizations

While many Azure Machine Learning and Cognitive Services applications are focused on business and consumer services, they’re also important tools for governments and other public bodies. Machine learning–powered tools can help make organizations like these more efficient, removing bottlenecks and speeding up common processes. Microsoft has been championing these approaches, with initiatives like its AI for Good program.

One important role for Microsoft’s AI tools is in breaking down barriers between different nationalities by providing tools for rapid, automatic translation. If you’ve used the captioning tools in PowerPoint or the Translator app on your smartphone, you’re using tools built around Azure’s speech recognition and translation services. We’ve looked at how to use them in your applications in Chapter 4, showing how speech recognition tools convert speech to text, how translation tools turn that text from one language to another, and how neural speech models deliver natural-sounding speech from translated text.

Delivering Translations for an International Parliament

Much of what we do with these tools is used to support individuals, translating menus or helping take a taxi across an unfamiliar city. But what if we need to provide near-real-time transcriptions of a large number of people, working in multiple languages, using a specialized vocabulary? That was the problem a team of Microsoft ...