Summary

In this chapter, we saw how we can create an app that produces captions in real time for a camera feed using deep CNNs and LSTMs. We also saw how we can quickly deploy some machine learning/deep learning models present in the form of Docker images to Red Hat OpenShift and easily obtain them in the form of callable APIs. This is very crucial from the perspective of an application developer as, when working with a team of machine learning developers, they will often provide you with Docker images of models to work with, such that you are not required to perform any code or configuration on the system. Such applications can be put to several uses, such as creating assistive technology for blind people, generating transcripts of events happening at that moment, or—say—having a live tutor for children to help them identify the objects in their environment...