Skip to content
👦 Basic face landmarking on iPhone with Dlib via Swift & ObjC++
C++ CMake C Swift Makefile Objective-C++
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.

Face Landmarking on iPhone

This prototype shows basic face landmark recognition on a CMSampleBuffer (see coming out of an AVCaptureSession.

Frame rate is actually quite good on an iPhone 6S now that we are using the system face detection via AVCaptureMetadataOutput. I did not measure performance yet but there is no discernible lag anymore. It looks like 30fps.

But I am sure there are a lot more performance improvements to be made. Currently, the buffers are copied around a lot.




This app uses the Dlib library ( and their default face landmarking model file downloaded from Thanks for the great work.

This project includes a precompiled Dlib. If you want to change something, consider that the Preprocessor Macros in the project linking Dlib need to be the same as the Compiler Flags when building the lib.

The project to build Dlib on iOS was generated according to these instructions.

Thanks to Satya Mallick from He recommended using the system face detector to me.


Code (except for DisplayLiveSamples/lib/*) is released under MIT license.

You can’t perform that action at this time.