In this modern age, sending data to AI-based applications in the cloud for inference is commonplace. For instance, a user can send an image taken on a mobile phone to the Amazon Rekognition API, and the service can tag the various objects, people, text, scenes, and so on, present in the image. The advantage of employing the service of an AI-based application that's hosted in the cloud is its ease of use. The mobile app just needs to make an HTTPS request to the AI-based service, along with the image, and, within a few seconds, the service will provide the inference results. A few of these machine learning as a service providers are as follows:
- Amazon Rekognition
- Amazon Polly
- Amazon Lex
- Microsoft Azure Cognitive Services
- IBM Watson
- Google Cloud Vision
The following diagram, Figure 7.1, illustrates the architecture of this kind...
