Check out this great blog post from Adrian Rosebrock at PyImageSearch.com on how to use Intel’s Movidius Neural Compute Stick with the Raspberry Pi!  This is a fascinating demonstration of applying modern machine learning and artificial intelligence to Python-powered hardware like the Raspberry Pi.  You might actually have already seen a demonstration like this with the Raspberry Pi Vision Kit, another bit of Pi hardware that uses the same Movidius neural network accelerator.  However Adrian shows that you can use a USB stick version of the accelerator to get very similar results as the vision kit.  Specifically you can perform real-time video object detection and much more using the powerful neural network co-processor connected to the Pi’s USB ports.  The post goes into great detail about some quirks of working with the hardware, and how to setup Python and OpenCV to process data with its models.  Check it out for more details:

What is the Intel Movidius Neural Compute Stick?

Intel’s Neural Compute Stick is a USB-thumb-drive-sized deep learning machine.

You can think of the NCS like a USB powered GPU, although that is quite the overstatement — it is not a GPU, and it can only be used for prediction/inference, not training.

I would actually classify the NCS as a coprocessor. It’s got one purpose: running (forward-pass) neural network calculations. In our case, we’ll be using the NCS for image classification.

The NCS should not be used for training a neural network model, rather it is designed for deployable models. Since the device is meant to be used on single board computers such as the Raspberry Pi, the power draw is meant to be minimal, making it inappropriate for actually training a network.

…

Under the hood of the NCS is a Myriad 2 processor (28 nm) capable of 80-150 GFLOPS performance. This processor is also known as a Vision Processing Unit (or vision accelerator) and it consumes only 1W of power (for reference, the Raspberry Pi 3 B consumes 1.2W with HDMI off, LEDs off, and WiFi on).