This is a read image and predict tflite model. Running Tensorflow Lite on iOS, Android, MacOS, Windows and Linux using Python and Kivy.
Create or use any tflite model.
- Check Neural Network input data.
- You can find out what size the neural network wants to input using model.py/get_input_shape().
pip3 install --upgrade pip
pip3 install -r requipments.txt
We use adb for logging android python executing commands.
python3 main.py
buildozer init
buildozer -v android debug deploy run
adb logcat | grep python
- Don't Change any things on .spec file
source.include_exts = py,png,jpg,kv,atlas,tflite
android.gradle_dependencies = 'org.tensorflow:tensorflow-lite:1.12.0','org.tensorflow:tensorflow-lite-support:0.0.0-nightly-SNAPSHOT'
requirements = python3,kivy,numpy,opencv
android.arch = arm64-v8a, armeabi-v7a
https://www.tensorflow.org/lite/models/modify/model_maker/object_detection