An open source IoT stack that utilizing computer vision and facilitating scalable interaction on the edge based on user pose, built with node.js, tensorflow, Rpi3 and FiPy
PoseBot is a stack that facilitates scalable cognitive interaction with actuators on the edge. The pose estimation model is executing on the client’s side and triggers actions to the end devices through a local web server.
- connect rpi to ethernet with internet
- connect rpi to power
- connect to posenet hotspot with pass posenetS
- connect fipys to power
- visit https://10.42.0.1:8080/
- play
PoseBot developed as an emerge of utilizing front-end machine learning techniques along with end devices in a proof of concept application. Both the width of the pose estimation algorithm and the reactions of the end devices are kept in a simple level, with the sole aim of represent the potentiality.