Face Recognition

• ROS package that applies face_recognition
• Tested Kinect V1 RGB and usb_cam
• For voice commands one must go to the repository and configure tts.

Installation

Building

cd catkin_ws/src
git clone https://github.com/UtBotsAtHome-UTFPR/utbots_face_recognition.git
cd ..
catkin_make

Dependencies

This package depends on freenect_launch or usb_cam and runs on python, as well as some official ROS dependencies.

The code runs on Python 3.8 and you must use a virtualenv (Install with pip install virtualenv) with the path /usr/bin/venv_utbots_face_recognition/bin/python as the node expects its existence to run. Install the requirements:

cd /usr/bin
sudo python3 -m virtualenv venv_utbots_face_recognition --python=$(which python3)
roscd utbots_face_recognition/src
/usr/bin/venv_utbots_face_recognition/bin/python -m pip install -r requirements.txt
cd
sudo rosdep init
rosdep install --from-paths src --ignore-src -r -y

For utbots voice one must go to the repository and configure tts.

Running

First, to run on usb:

roslaunch utbots_face_recognition usb_action_server.launch

To run on kinect:

roslaunch utbots_face_recognition freenect_action_server.launch

Actions

new_face_action
train.action
recognition.action