Skip to content
An open source library for face detection in images. The face detection speed can reach 1500FPS.
Branch: master
Clone or download
ShiqiYu a bug fixed
A memory violation bug fixed.
Latest commit 54b8e03 Sep 24, 2019
Type Name Latest commit message Commit time
Failed to load latest commit information.
example Update the comment on the input image format Mar 28, 2019
images Update the example image. Oct 23, 2018
mobile update apk for bgr Mar 28, 2019
models The model is updated to v2. Sep 14, 2019
src a bug fixed Sep 24, 2019
CMakeLists.txt The model is updated to v2. Sep 14, 2019
ChangeLog The model is updated to v2. Sep 14, 2019
LICENSE Open Source! Mar 14, 2019 The model is updated to v2. Sep 14, 2019
aarch64-toolchain.cmake build static and shared libraries Mar 15, 2019


This is an open source library for CNN-based face detection in images. The CNN model has been converted to static variables in C source files. The source code does not depend on any other libraries. What you need is just a C++ compiler. You can compile the source code under Windows, Linux, ARM and any platform with a C++ compiler.

SIMD instructions are used to speed up the detection. You can enable AVX2 if you use Intel CPU or NEON for ARM.

The model file has also been provided in directory ./models/.

examples/libfacedetectcnn-example.cpp shows how to use the library.


How to use the code

You can copy the files in directory src/ into your project, and compile them as the other files in your project. The source code is written in standard C/C++. It should be compiled at any platform which support C/C++.

Some tips:

  • Please add -O3 to turn on optimizations when you compile the source code using g++.
  • Please choose 'Maximize Speed/-O2' when you compile the source code using Microsoft Visual Studio.
  • ENABLE_INT8=ON is recommended for ARM, but it is not recommended for Intel CPU since it cannot gain better speed sometime even worse.
  • The source code can only run in single thread. If you want to run parally, you can call the face detection function in multiple threads. Yes, multiple-thread is complex in programming.
  • If you want to achieve best performance, you can run the model (not the source code) using OpenVINO on Intel CPU or Tengine on ARM CPU.

If you want to compile and run the example, you can create a build folder first, then run the command:

mkdir build; cd build; rm -rf *

Use Tengine to Speedup the detection on ARM

The model has been added to Tengine. Tengine, developed by OPEN AI LAB, is a lite, high-performance, and modular inference engine for embedded device.

The model in Tengine can run faster than the C++ source code here because Tengine has been optimized according to ARM CPU. There are detailed manual and example at Tengine web site:

Cross build for aarch64

  1. Set cross compiler for aarch64 (please refer to aarch64-toolchain.cmake)
  2. Set opencv path since the example code depends on opencv
cmake \
    -DCMAKE_TOOLCHAIN_FILE=../aarch64-toolchain.cmake \


Native build for avx2

cmake \
    -DDEMO=ON \


CNN-based Face Detection on Windows

Method Time FPS Time FPS
X64 X64 X64 X64
Single-thread Single-thread Multi-thread Multi-thread
OpenCV Haar+AdaBoost (640x480) -- -- 12.33ms 81.1
cnn (CPU, 640x480) 64.55ms 15.49 15.78ms 63.36
cnn (CPU, 320x240) 15.48ms 64.60 3.92ms 255.01
cnn (CPU, 160x120) 3.86ms 259.01 1.07ms 938.71
cnn (CPU, 128x96) 2.46ms 406.33 0.68ms 1479.79
  • OpenCV Haar+AdaBoost runs with minimal face size 48x48
  • Face detection only, and no landmark detection included
  • Minimal face size ~10x10
  • Intel(R) Core(TM) i7-7700 CPU @ 3.6GHz

CNN-based Face Detection on ARM Linux (Raspberry Pi 3 B+)

(* to be updated) | Method |Time | FPS |Time | FPS | |--------------------|--------------|-------------|--------------|-------------| | |Single-thread |Single-thread|Multi-thread |Multi-thread | |cnn (CPU, 640x480) | 512.04ms | 1.95 | 174.89ms | 5.72 | |cnn (CPU, 320x240) | 123.47ms | 8.10 | 42.13ms | 23.74 | |cnn (CPU, 160x120) | 27.42ms | 36.47 | 9.75ms | 102.58 | |cnn (CPU, 128x96) | 17.78ms | 56.24 | 6.12ms | 163.50 |

  • Face detection only, and no landmark detection included.
  • Minimal face size ~10x10
  • Raspberry Pi 3 B+, Broadcom BCM2837B0, Cortex-A53 (ARMv8) 64-bit SoC @ 1.4GHz



Some contributors are listed here.

The contributors who are not listed at

  • Jia Wu (吴佳)
  • Dong Xu (徐栋)
  • Shengyin Wu (伍圣寅)


The work is partly supported by the Science Foundation of Shenzhen (Grant No. JCYJ20150324141711699 and 20170504160426188).

You can’t perform that action at this time.