Qt TensorFlow Lite example

A simple example that shows how to use TensorFlow Lite with Qt. That’s a modified version of the label_image example.

Tested on Ubuntu 20.04.

How to build

• Install Visual Studio Code with Remote Containers extension
• Clone this project and open in Visual Studio Code

   git clone https://github.com/IvanSafonov/qt-tf-lite-example.git
   code ./qt-tf-lite-example

• In the Command Palette find and click Remote-Containers: Reopen in Container
• In the Command Palette find and click CMake: Build

Launch example

The result executable file can be found in the build subdirectory.

vscode ➜ /workspaces/qt-tf-lite-example $ ./build/qt-tf-lite-example -h
Usage: ./build/qt-tf-lite-example [options] images
Qt TensorFlow Lite example
Options:
  -h, --help             Displays this help.
  -m, --model <model>    TFLite model
  -l, --labels <labels>  Model labels file
Arguments:
  images                 Input images for classification

vscode ➜ /workspaces/qt-tf-lite-example $ ./build/qt-tf-lite-example ./grace_hopper.jpg
INFO: Initialized TensorFlow Lite runtime.
"./grace_hopper.jpg"
0.784314 653 "military uniform"
0.105882 907 "Windsor tie"
0.0156863 458 "bow tie"
0.0117647 466 "bulletproof vest"