Google’s AIY Projects, a clever play on“AI” and“DIY”, comprise a Raspberry Pi with the parts, essential cardboard, and software to make your own object-recognizing smart camera or Google Assistant smart speaker. They only rolled out last year, but Google already launches new and upgraded versions of both the AIY Voice Kit and AIY Vision Kit recently.
The new kits now include a preloaded SD card and a new Raspberry Pi Zero in the box to make it easier to operate. Now, Google is also offering an AIY companion app for Android (and soon Chrome and iOS) to help you control configuration and wireless setup in your new DIY smart camera or speaker. Although you can still use the Pi as a standard Linux computer with a monitor, mouse, and keyboard if that’s more your style. The Vision Kit also includes the Pi Camera Module V2.
Google is working in a lot of areas to make it easier to across machine learning. You can download the pre-trained models for your Vision Kit from the AIY Projects site. Besides, Google just launched TensorFlow.js, which is a version of TensorFlow that based on browser, Google’s industry-leading machine learning framework. You can image that now you can train your own cat-recognizer in a web browser and then run it on your self-built smart camera without having a PhD in machine vision.