This is the blog of Matthijs Hollemans about machine learning and AI.
Buy my e-book Core ML Survival Guide
This guide is a massive collection of tips and tricks for working with Core ML and mlmodel files. For simple tasks Core ML is very easy to use… but what do you do when Core ML is giving you trouble? The solution is most likely in this 400+ page book! It contains pretty much everything I learned about Core ML over the past few years. Check it out at Leanpub.com
Hire me to add AI to your app
Want to make your iOS apps smarter? I can help you choose a machine learning model and integrate it into your app. For apps that are pushing the envelope, I can design and train custom models. If you already have a model, I can optimize it for use on mobile. Read about my services
11 Jan 2021
— Implementation of the SEFR classification algorithm in Swift
25 Nov 2020
— What to do if the predicted bounding boxes are drawn in the wrong place
29 Jun 2020
— A look at what has changed in Apple’s machine learning APIs for iOS and macOS
8 Apr 2020
— Investigating which model architecture makes the best backbone on iPhone
6 Feb 2020
— Because bilinear interpolation does not work the same everywhere
9 Dec 2019
— Make your Core ML models work directly with images instead of multi-arrays
1 Dec 2019
— How to use the new on-device model personalization APIs from Core ML 3
13 Nov 2019
— Using Core ML with Apple’s new reactive programming framework
14 Sep 2019
— How to use the new on-device model personalization APIs from Core ML 3
10 Aug 2019
— How to use the new on-device model personalization APIs from Core ML 3
19 Jul 2019
— How to use the new on-device model personalization APIs from Core ML 3
8 Jun 2019
— What is new in Core ML 3, including all the new neural network layer types
17 Dec 2018
— How to convert SSD to work with Vision’s new object detection API
5 Dec 2018
— Announcing two new books: Machine Learning by Tutorials and the Core ML Survival Guide
16 Aug 2018
— Some examples of how algorithms and data structures are used in real iOS apps
30 Jun 2018
— Finding out how much computation and memory is needed by deep neural networks
9 Jun 2018
— An in-depth look at how fast object detection models are trained
22 Apr 2018
— What is new in MobileNet version 2, and how it stacks up against V1
11 Dec 2017
— How to create your own custom layers with Core ML neural networks
22 Nov 2017
— Machine learning on mobile is currently inference only, but what about training?
2 Sep 2017
— Making a deep convolutional neural network smaller and faster
21 Aug 2017
— Reverse engineering Core ML and the compute kernels it uses under the hood
26 Jul 2017
— What to do when your Core ML model does not output the results you were expecting
23 Jul 2017
— Core ML, Metal Performance Shaders, TensorFlow, or roll your own?
21 Jun 2017
— Showdown between the two new Apple machine learning APIs
14 Jun 2017
— Implementing the MobileNet architecture on iOS. Is it really as fast as its inventors claim?
11 Jun 2017
— Overview and opinions about the new machine learning APIs announced for iOS 11
20 May 2017
— Implementing the YOLO object detection neural network in Metal on iOS
24 Apr 2017
— An open source library that makes it easy to build neural networks with MPSCNN
6 Apr 2017
— Using an LSTM to teach the iPhone how to play the drums
6 Mar 2017
— Learn how to build machine learning models with TensorFlow and put them into your apps
22 Feb 2017
— Comparing matrix multiplication on the GPU against BLAS and fully-connected layers
16 Feb 2017
— The choices to make when you decide to add deep learning to your mobile app
7 Feb 2017
— Comparing the two deep learning APIs Apple introduced in iOS 10
18 Jan 2017
— Learn how OpenGL and Metal work by writing your own 3D renderer from scratch
30 Aug 2016
— Building an image recognition app using Metal
24 Aug 2016
— Using Apple’s new BNNS framework to make a basic neural network
25 Mar 2016
— Take advantage of Swift’s type system to make your programs more expressive and avoid silly mistakes
22 Jul 2015
— A transcript of my talk at the Dutch CocoaHeads meetup in Rotterdam