CoreSense4Home examples

This repository contains small ROS 2 examples built with cs4home_architecture. The examples show how sensor information is received by a cognitive module, processed by its Core component and published for use by the rest of the system.

Two examples are included:

• Face identities from hri_face_detect are converted into knowledge-graph updates.
• Camera images are processed by the YOLO example and published as detections.

CoreSense role

The terms below follow the CoreSense Ontology (CSO).

• A camera or face tracker supplies observations originating from a Sensor.
• Each example implements a Cognitive Function that transforms incoming information.
• The YOLO example contributes a visual-perception Cognitive Capability.
• The face example contributes information that can be retained by a World Model.

Example flow

flowchart LR
    faces["Tracked face identities"] --> face["Face cognitive function"]
    face --> knowledge_graph["Knowledge-graph update"]
    camera["Camera image"] --> yolo["YOLO cognitive function"]
    yolo --> detections["Object detections"]

Loading

Build

mkdir -p ~/cs4home_examples_ws/src
cd ~/cs4home_examples_ws/src
git clone https://github.com/CoreSenseEU/cs4home_examples.git
vcs import --recursive < cs4home_examples/thirdparty.repos
cd ..
python3 -m venv --system-site-packages py_deps
source py_deps/bin/activate
pip install -r src/thirdparty/hri_face_detect/requirements.txt
pip install -r src/thirdparty/yolov8_ros/requirements.txt
rosdep install --from-paths src --ignore-src -r -y
colcon build --symlink-install
source install/setup.bash

Run

Start the camera and YOLO dependency, then run:

ros2 launch cs4home_simple_project yolo_example.launch.py

For the face example, start hri_face_detect and run the face module with the supplied parameters:

ros2 run cs4home_simple_project face_module --ros-args \
  --params-file src/cs4home_examples/cs4home_simple_project/config/params_simple.yaml

The face result can be inspected with the knowledge-graph RQt plugin.