Peg base openvino image to 2021.4
Parameterize hard-coded values

Change-Id: I0c3d4866709b5ea98056699818fbac53da979af4
7 files changed
tree: fa51b8013070f6a0986e39bfdc7e099fffc371c8
  1. .gitignore
  2. LICENSES/
  3. Makefile
  4. README.md
  5. VERSION
  6. bin/
  7. camera/
  8. docker/
  9. models/
  10. person_detection/
  11. requirements.txt
  12. resources/
README.md

Person-detection Application

Person-detection-app is a demo application that runs in the Aether edge and showcases Aether's support for for low-latency, AI/ML driven IOT applications that make use of Aether's support for end-to-end slicing.

The person-detection-app consists of a client component that runs on Raspberry-Pi camera and a server component that runs on a server in the Aether Edge.

Getting started

Raspberry Pi Camera

The camera directory contains scripts that run on the Raspberry Pi camera.

  • openvino-camera.sh captures video from the Raspberry Pi's camera module, encodes and transmits it in low or high resolution to the host:port of the person-detection-app
  • openvino-mqtt.py listens to mqtt commands from the person-detection-app to set the resolution level of the Raspberry Pi camera video.

It is recommended that these two scripts be started by systemd on startup.

Person-detection-app

The person-detection app runs as a docker container.

docker pull registry.aetherproject.org/aether-apps/person-detection-app:latest
docker run -itu root:root --privileged --network host registry.aetherproject.org/aether-apps/person-detection-app:latest