https://github.com/seanchas116/tflitesegmentationqt

Run TensorFlow Lite image segmentation in Qt desktop app
https://github.com/seanchas116/tflitesegmentationqt

cpp qt qt-desktop tensorflow tensorflow-lite

Last synced: about 1 month ago
JSON representation

Run TensorFlow Lite image segmentation in Qt desktop app

• Host: GitHub
• URL: https://github.com/seanchas116/tflitesegmentationqt
• Owner: seanchas116
• License: mit
• Created: 2019-10-22T06:50:49.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2019-10-25T06:49:24.000Z (over 5 years ago)
• Last Synced: 2024-11-08T20:50:23.778Z (3 months ago)
• Topics: cpp, qt, qt-desktop, tensorflow, tensorflow-lite
• Language: C++
• Homepage:
• Size: 26.6 MB
• Stars: 4
• Watchers: 3
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# TFLiteSegmentationQt

Run TensorFlow Lite image segmentation in Qt desktop app (currenly for Mac only)

![image](screenshot.png)

## How to run

Open `CMakeLists.txt` with Qt Creator, or build the project with cmake manually.

## Install tensorflow-lite

This repo already includes tensorflow-lite build (v2.0.0), but if you want to use tensorflow-lite built by yourself do the following:

* Clone [tensorflow repo](https://github.com/tensorflow/tensorflow)

* Build tensorflow-lite for desktop

  * If you are using Windows, use MSYS2

```

cd path/to/tensorflow

./tensorflow/lite/tools/make/download_dependencies.sh

./tensorflow/lite/tools/make/build_lib.sh

```

* Copy .a and .h

```

# Copy TensorFlow Lite headers

cd path/to/tensorflow/tensorflow/lite

gcp --parents **/*.h /path/to/tflite-qt-app/vendor/tensorflow-lite/include/tensorflow/lite

# Copy flatbuffers headers

cp -r tools/make/downloads/flatbuffers/include/flatbuffers /path/to/tflite-qt-app/vendor/tensorflow-lite/include

# Copy libtensorflow-lite.a (platform name may vary)

cp tools/make/gen/osx_x86_64/lib/libtensorflow-lite.a /path/to/tflite-qt-app/vendor/tensorflow-lite/osx_x86_64/lib

```