Object detector architecture

We will be recreating the Amazon Rekognition demo with our own web frontend and Python backend. First, let's understand the architecture of the Object Detector application we are about to develop.

Using the reference architecture design and the technology stack we discussed previously, here is the architecture for the Object Detector application:

The user will interact with the Object Detector through a web user interface:

• We will provide a web user interface for users so that they can see the Object Detection demo.
• The web user interface will interact with the orchestration layer containing just one RESTful endpoint: the Demo Object Detection endpoint.
• The endpoint interacts with both the Storage Service and the Recognition Service to perform the Object Detection demo.
• The Storage Service and the Recognition Service calls the Amazon S3 and Amazon Rekognition services using the Boto3 SDK, respectively.