Mobile Deep Learning with TensorFlow Lite, ML Kit and Flutter

Generating image captions from the camera feed

Now that we have a clear idea about the image caption generator and have an application with a camera feed, we are ready to generate captions for the images from the camera feed. The logic to be followed is very simple. Images are captured from the live camera feed at a specific time interval and are stored in the device's local storage. Next, the stored pictures are retrieved and an HTTP POST request is created for the hosted model, passing in the retrieved image to fetch the generated captions, parsing the response, and displaying it on the screen.

Let's now look at the detailed steps, as follows: 

1. We will first add an http dependency to the pubspec.yaml file to make http requests, as follows: ...