insightface gender age model dynamic lib (Python callable)

How to use

download the required 3rdparty dynamic lib folder from link, extraction code-d3hh put it in project's root path, run docker build -t '$IMAGENAME:TAG' . for building opencv3.4, python2.7, gcc and cmake etc environment and then run bash ./run.sh will create a shared lib - ./build/libexport_gender_age.so and run python demo - predict.py at the same time.

Config file

the config file: config.ini, you can change the file paths etc here

Logger

modify the [LOG]->LEVEL in the config.ini to change logging.

enum LoggerLevel
{
    none = 0,
    fatal = 1,
    error = 2,
    warning = 3,
    info = 4,
    debug = 5,
    verbose = 6
};

Reference

[1] insightface: https://github.com/deepinsight/insightface
[2] mxnet c++ inference example: https://mxnet.apache.org/versions/1.6/api/cpp/docs/tutorials/cpp_inference.html
[3] extract feature from pretrained model by mxnet c++ api: https://blog.csdn.net/muyouhang/article/details/85059352
[4] build opencv from source in ubtunu16.04: https://www.learnopencv.com/install-opencv-4-on-ubuntu-16-04/
[5] dockerfile source: https://hub.docker.com/r/dymat/opencv/dockerfile