It has arrived. With iOS 11, Apple finally introduced a native machine learning and machine vision framework. This opens up a whole host of new possibilities, promising great leaps forward in apps and games of all natures.
Machine learning solutions have been available for a while in the cloud, but these systems require a constant internet connection and oftentimes have a very noticiable delay on iOS for obvious reasons. This also creates a security risk for sensitive data. Some third-party Swift AI systems have begun to take hold inside of a select few apps, but such frameworks never hit the mainstream development community. With Apple’s new introduction at the 2017 WWDC it is likely that many of your favorite apps will see signifigant machine-learning-related updates.
You can find it in here if you are interested in seeing how you can integrate Apple’s new APIs into your very own apps.
It’s easier than you think!
- Xcode 9
- Swift 4
- iOS 11
To run this project, you need a pre-trained model. I have put a model in this project and it is one of the best as reference accuracy and size of the model. However, you can choose another model in here what you want.
Clone it with:
git clone git@github.com:Wanguy/Core-ML-and-Inception-v3.git
cd Core-ML-and-Inception-v3
or manually download.
Copyright (c) 2015-2017 Wanguy. Released under GPLv3. See LICENSE for details.