Link Search Menu Expand Document

VOXL-TFLite-Server

The voxl-tflite-server is a GPU-optimized TensorFlow Lite envirnoment. It processes standard TensorFlow Lite models in real-time using the GPU on VOXL.

You can find the code and README on Gitlab

Configuration

By default, the voxl-tflite-server reads images from voxl-camera-server on the /run/mpa/hires_preview/ pipe

The config file can be found /etc/modalai/voxl-tflite-server.conf

Running

The service can be configured to run automatically on boot by following these steps on voxl (via adb or ssh):

yocto:/$ systemctl enable voxl-tflite-server
yocto:/$ systemctl start voxl-tflite-server

To validate that the service is now running, run:

yocto:/$ voxl-inspect-services

Make sure that voxl-tflite-server is both enabled and running.

Viewing

The output of voxl-tflite-server can be viewed using voxl-streamer

Demo