Modern Deep Learning Module

The module is wrapper to tiny-dnn, a header only, dependency-free deep learning framework in C++11.

Installation

Required Dependencies

• System under Unix or Windows
• C++11 compiler
• tiny-dnn headers
• protoc compiler (part of the protobuf library)

How to install tiny-dnn? CMake will try to download a certain version of tiny-dnn which was tested to build with OpenCV. If the latest version is needed, location of the tiny-dnn folder can be specified manually:

• Download tiny-dnn project somewhere in your system

    cd /opt
    git clone https://github.com/tiny-dnn/tiny-dnn.git

• Run your OpenCV CMake pointing to your tiny-dnn headers location

    cd /opt/opencv/build
    cmake -DTINYDNN_ROOT=/opt/tiny-dnn ..
    make -j4

Extra

You can enable some optimizations just for tiny-dnn backend

cmake -DTINYDNN_USE_SSE=ON ..
cmake -DTINYDNN_USE_AVX=ON ..

Use third-party multithreading libs: TBB or OMP.

cmake -DTINYDNN_USE_TBB=ON ..     // then disable OMP
cmake -DTINYDNN_USE_OMP=ON ..     // then disable TBB

NNPACK: Acceleration package for neural networks on multi-core CPUs.
Check project site for installation: https://github.com/Maratyszcza/NNPACK

cmake -DTINYDNN_USE_NNPACK=ON ..  // not supported yet for Caffe loader

See detailed module API documentation in http://docs.opencv.org/trunk/d1/df7/group__dnn__modern.html