RPI4 Google Coral USB Performance

[cigas4]

Hello,

I am running Frigate container on RPI4 together with 3 IP cameras. To improve performance I have ordered and installed Google Coral USB accelerator, but for some reason I do not see any inprovement or a dropdown of cpu load on my rpi4..
Coral seems to be recognized by the frigate but I dont see any efect when it is there, compared to bare rpi4.
Do I need to make some configurations or drivers setup on rpi to use coral?

My docker compose config:

version: '3.9'
services:
  frigate:
    container_name: frigate
    privileged: true
    restart: unless-stopped
    image: blakeblackshear/frigate:stable-armv7
    devices:
      - /dev/bus/usb:/dev/bus/usb  
    volumes:
      - /etc/localtime:/etc/localtime:ro
      - /home/pi/Downloads/frigate/frigate.yml:/config/config.yml:ro
      - /media/frigate:/media/frigate
      - type: tmpfs # Optional: 1GB of memory, reduces SSD/SD Card wear
        target: /tmp/cache
        tmpfs:
          size: 1000000000
    ports:
      - '5000:5000'
      - '1935:1935'
    environment:
      FRIGATE_RTSP_PASSWORD: 'pass'

See CPU uage graphs for both using cpu and coral detectors, cpu usage looks almost identical.

Coral USB detector

RPI4 CPU1&CPU2 detector

[blakeblackshear]

The Coral is only used when object detection runs, so you won't see any change in CPU usage for decoding the video streams. When tracking an object you should see very little increase in CPU usage when using the Coral compared to CPU detectors.