roboflow_general

roboflow general

Runs roboflow object-detection and instance-segmentation models.

Additional installation

None.

Usage

First add a camera device to MeetingCam

To show a modified camera stream in online meetings it is necessary to create a virtual camera for each real camera you want to use.

1. Activate the virtual environment

   source .venv/bin/activate

2. Get command to add a camera device

   python src/meetingcam/main.py add-devices

   Then copy and execute the command of the camera you want to use.
3. You can now see that the camera of your choice has a virtual counterpart.

   python src/meetingcam/main.py list-devices

Run a plugin

1. Currently it is needed to run a roboflow inference server in a separate terminal. To do so, simply run the following to start an inference server (CPU).

   docker run --net=host roboflow/roboflow-inference-server-cpu:latest
   

   See here for more options.

2. Run the plugin with replacements for api-key, project-name, version and device_path with your camera path which has a virtual counterpart.

   python src/meetingcam/main.py roboflow --api-key YOUR_API_KEY --project-name meetingcam-roboflow-example-objectdetection --version 1 device_path

   As an alternative you can run the following and you'll be ask to enter the above replacements by terminal prompt.

   python src/meetingcam/main.py roboflow /dev/video0

You can hide the label imprints by pressing <Ctrl+Alt+l>.

Get help

Options and usage for the plugin are documented in the plugins help function.

python src/meetingcam/main.py roboflow

Showcase

Example - Object Detection

This is a custom roboflow project which runs on MeetingCam like below.

Roboflow project page here.

Example - Instance Segmentation

This is a custom roboflow project which runs on MeetingCam like below.

Roboflow project page here.

Implementation details

Currently just object-detection and instance-segmentation models are supported.