Google's data centres hold trillions of terabytes of data, from media querries to images, google has it all in abundance. If you are familiar with supervised learning, you will know that on such a large stack of data there rests a great deal of machine learning potential. Unfortunately, the average data scientist (me) doesn't have access to this glut of data, and even if I did, I doubt my laptop could handle it. Alas. Yet there is some good news, software developers, you are in luck! Google has been using its vast data stores to do plenty of machine learning for you!
Google's Firebase, Google's app development platform, recently began offering a range of machine learning APIs for easy integration of advanced machine learning techniques. Image labelling? Optical Character Recognition? They have you covered! It is exciting stuff, I am sure you agree. So exciting that I had to see it working for myself. Introducing SmartList!
The idea is straightforward and potentially useful. Have you ever been reading a cookbook and needed to convert an ingredients list to a convenient shopping list on your phone? Yes, I thought so. Well! Worry no more! With this simple mobile application (only on android), your dreams are now a reality! Simply utilize your smart device's camera to capture an image of the aforementioned ingredients list; do a quick bit of manual cropping; then watch as computer magic creates your shopping list! Check it out now! (and check out the source code on GitHub)!
That spiel aside, I am genuinely impressed by this little app. It works well, and it was put together in just a couple of workdays. It is quite mind-boggling when you think about it. Firebase can analyse an entire image in (nearly) real-time, it can read faster than people!
Download it here: https://play.google.com/store/apps/details?id=com.adam.shoppinglist
And see the source code here: https://github.com/adamprice97/SmartList