DaisyKit Documentation  dev-0.1.x
AI framework focusing on the ease of deployment

Daisykit is an AI toolkit for software engineers to Deploy AI Systems Yourself (DAISY). We develop this package with a focus on the ease of deployment. This repository contains:

  • Daisykit SDK, the core of models and algorithms.
  • Daisykit Python.

Website: https://daisykit.org/.

Demo Video: https://www.youtube.com/watch?v=zKP8sgGoFMc.


DaisyKit Architecture

Our development plan for Daisykit. We are working to build the whole system gradually.

Environment Setup

For Ubuntu, we need build tools from build-essential package. For Windows, Visual Studio 2019 is recommended.

  • Install OpenCV.


sudo apt install libopencv-dev


Download and extract OpenCV from the official website, and add OpenCV_DIR to path.

  • Install Vulkan development package.


sudo apt install -y libvulkan-dev vulkan-utils
sudo apt install mesa-vulkan-drivers # For Intel GPU support
  • Download precompiled NCNN, extract it (version for your development computer).

Build and Run on PC

  • Initialize / Update submodules
git submodule update --init
  • Build


mkdir build
cd build
cmake .. -Dncnn_FIND_PATH="<path to ncnn lib>"


mkdir build
cd build
cmake -G "Visual Studio 16 2019" -Dncnn_FIND_PATH="<path to ncnn lib>" ..
cmake --build . --config Release
  • Run face detection example





Coding convention

Read coding convention and contribution guidelines here.

Build documentation

  • Step 1: Install doxygen first.
  • Step 2: Build the documentation:
cd docs
doxygen Doxyfile.in
  • Step 3: Deploy html documentation from docs/_build/html.
  • Step 4: Our lastest documentation is deployed at https://docs.daisykit.org.

Known issues and problems

1. Slow model inference - Low FPS

This issue can happen on development build. Add -DCMAKE_BUILD_TYPE=Debug to cmake command and build again. The FPS can be much better.


This toolkit is developed on top of other source code. Including