If you are one of those people who likes to use your Raspberry Pi for much more than connecting it to your television and playing multimedia content, you will surely like to know the second project of Google that was born within the Artificial Intelligence Yourserf initiative, same for which the collaboration of the Raspberry Pi Foundation itself.
The idea behind this Google initiative is to demonstrate to all users how simple and interesting it can be to integrate TensorFlow, Google's artificial intelligence platform, in any project to enhance all its possibilities.
Google and Raspberry Pi united by a very special and interesting project for every maker
Returning to the project that they have just launched, tell you that it has been baptized with the name of Vision Kit and it is nothing other than a kind of image recognition camera that does not need to be connected to any type of web service or server to perform different advanced functions because it has its own neural network model.
So that you can assemble this smart camera yourself, you will have to buy a kit composed of different objects such as a Raspberry Pi Zero W, a Raspberry Pi Camera 2 that, together with the controller, you must connect to a VisionBonnet board, lenses, cables and a cardboard box to put all this hardware inside.
Personally I have to confess that the most interesting part of this project is found in the 'Vision Kit SD Image', a file that we must download and insert on an SD card and that includes nothing less than three neural networks based on TensorFlow. If you are interested in the kit, tell you that Google has decided to put it on sale at a price of U.S. dollar 44,99.