is working on recognizing playing cards so he first sprinkled them around his living room, added some clutter, and walked around, taking pictures using his phone. This may sound like a long, tedious task, but it can be done efficiently. Fill your background with various other things and even have some things partially obscuring your objects. If you use the latter approach, make sure to shoot from various angles, rotations, and with different lighting conditions. These can either be scraped from an online source like Google’s images or you get take your own photos. You’ll need a few hundred images of your objects. MobileNet is an example of one which is less accurate but recognizes faster and so is better for a Raspberry Pi or mobile phone. Inception is one of a few which are very accurate but it can take a few seconds to process each image and so is more suited to a fast laptop or desktop machine. You’re not limited to just Inception though. He does it for Windows 10 since there’s already plenty of documentation out there for Linux OSes. When the time comes to add an object recognizer to your hack, all you need do is choose from many of the available ones and retrain it for your particular objects of interest. To help with that, has put together a step-by-step guide to using TensorFlow to retrain Google’s Inception object recognizer.
