Detecting objects using TFHub

TFHub is a cornucopia of state-of-the-art models when it comes to object detection. As we'll discover in this recipe, using them to spot elements of interest in our images is a fairly straightforward task, especially considering they've been trained on the gigantic COCO dataset, which make them an excellent choice for out-of-the-box object detection.

Getting ready

First, we must install Pillow and TFHub, as follows:

$> pip install Pillow tensorflow-hub

Also, because some visualization tools we'll use live in the TensorFlow Object Detection API, we must install it. First, cd to a location of your preference and clone the tensorflow/models repository:

$> git clone –-depth 1 https://github.com/tensorflow/models

Next, install the TensorFlow Object Detection API, like this:

$> sudo apt install -y protobuf-compiler
$> cd models/research
$> protoc object_detection/protos/*.proto –-python_out=.
...