I believe on the website the YouTube video linked is using this library throughout the whole video to give you visual references based on what the person is talking about. It's a pretty good demo..
I second the need for samples. This looks great, but when there are no examples of it in use I have to wonder if it's really been tested in real use cases.