Need to implement the TensorFlow model Object Detection

So I have a chrome extension that can take a screenshot of any user-selected area and send it to Google Image Search. Therefore I don’t need to Google the Image by copy pasting the image-link nor download and then upload the image into Google Image search. Simply by taking a screenshot of a selected area in the chrome browser, I can search the taken screenshot in Google Image Search.

Now I need to implement the TensorFlow model Object Detection into my project. It should be able to tell me what I can see in the screenshot and it should be highlighted as well.
The TensorFlow Model I have to use is https://github.com/tensorflow/tfjs-models/tree/master/coco-ssd.
I need to implement them into my code (in a .html file) via script tags.