Perform inferencing of tensorflow-lite models on an RPi with acceleration from Coral USB stick
MIT License
Bot releases are hidden (Show)
Published by robmarkcole almost 2 years ago
Allow you to pass a min_confidence in the form post similar to how deepstack operates.
Published by robmarkcole over 2 years ago
Published by robmarkcole about 3 years ago
Update to use pycoral library
Published by robmarkcole over 4 years ago
Published by robmarkcole about 5 years ago
Adds basic logging to file and adds a service for running the app on restart of the pi
Published by robmarkcole about 5 years ago
Use args to allow setting:
Published by robmarkcole about 5 years ago
Use disk images from edgetpu-platforms to simplify setup for raspberry pi users
Published by robmarkcole about 5 years ago
Fixes an inconsistency with deepstack
Published by robmarkcole about 5 years ago
Returns data in same format as https://python.deepstack.cc/object-detection for use with https://github.com/robmarkcole/HASS-Deepstack-object
Published by robmarkcole over 5 years ago
Adds Docker file, big thanks to @lmamakos
Published by robmarkcole over 5 years ago
As title, see the jupyter notebook for example
Published by robmarkcole over 5 years ago
Returns a confidence in % (percentage)
Published by robmarkcole over 5 years ago
Basic functionality of loading the object detection model, and exposing predictions via Flask app