As a brief update on some of my project ideas for Nao (which I mentioned in my last blog post), I have prototyped access to Google data via the gData API. Using the Google Calendar as a test bed, Nao can access the secure, private calendar feed and read appointments out via his text to speech sub system. There is more information after the break but here is a short video.
Posted: 08/08/2011 21:32:56
I have been playing with some of the face recognition and camera features of Nao over the last few days, and I must confess it has been a lot of fun. Nao has the ability to recognise a human face using it's built in camera and also can take photographs. With the fact Nao is internet connect, we have been able to combine the features to create a little program that lets Nao take pictures of the people he meets and share them on the web.
Basically, when Nao recognises a face he politely asks if he can take a photograph for his website. If the person agrees (just by saying 'yes', voice recognition does the rest) Nao snaps a photograph and uploads it to his website. His website happens to be a page of this website so you can see the half a dozen test pictures we have taken so far here.
We were also quite surprised that he recognised a Gorilla's face and also a face in a photo.
A photo he took when meeting my brother is below:
Posted: 02/06/2011 21:47:20
Another great piece of Nao development from Caxisa. This time to achieve scoring with a ball through a basket ball hoop. The navigation is by locating and recognising a Nao Landmark above the basket. The Nao Landmark is recognised using one of the two video cameras in Nao's head and the size and position of the mark are used to determine distance and positioning. A great piece of work, thanks for sharing it with us.
Posted: 26/05/2011 22:04:45