5 Minute Quickstart from Hub

Getting Started

All models available from our hub come with an easy-to-use .cpp demo project.

Prerequities:

• cmake
• Visual Studio (on Windows)
• XCode with Clang (on Mac)

Steps

1. Download the package for your desired OS & hardware combination from github.com/blace-ai/blace-ai/releases.
2. Unpack the .zip file, you end up with a folder structure like

   package/
       - cmake
       - demo
       - include
       - lib / BlaceAI.app (on MacOS)
       - licenses

3. From the hub download the Demo Project and Pytorch Payload .bin file.
4. Unzip the demo project and put contained folder into the package folder. Additionally copy the downloaded .bin file into the ..._demo folder:

   package/
       - cmake
       - demo
       - <some_model_identifier>_demo/
           - <downloaded_model>.bin
           - CMakeLists.txt
           - ...
       - include
       - lib / BlaceAI.app (on MacOS)
       - licenses

5. In a terminal navigate into the demo folder: cd package/<some_model_identifier>_demo.
6. Invoke cmake configuration command:
   cmake -G "Visual Studio 17 2022" -A x64 -T v142 -S . -B build (Windows)
   cmake -S . -B build (MacOS)
   cmake -S . -B build (Linux)
   This creates project files in the build folder.
7. Invoke build command: cmake --build build --config Release
   This builds the demo executable and copies all needed libraries, the ai model file and an input image next to the executable.
8. Run the demo executable:
   .\build\Release\demo.exe (Windows)
   ./build/Release/demo (MacOS)
   ./build/demo (Linux)