/deepgaze_ros

ROS wrapper package for deepgaze

Primary LanguagePythonMIT LicenseMIT

deepgaze_ros

If you use any of this code please cite our paper "Human operator cognitive availability aware Mixed-Initiative control". You can find it in IEEE explore link and arxiv link.

G. Petousakis, M. Chiou, G. Nikolaou and R. Stolkin, "Human operator cognitive availability aware Mixed-Initiative control," 2020 IEEE International Conference on Human-Machine Systems (ICHMS), 2020, pp. 1-4, https://doi.org/10.1109/ICHMS49158.2020.9209582

ROS wrapper package for deepgaze

Install deepgaze library

For using the original deepgaze software cite Patacchiola et al as requested in their github page. https://github.com/mpatacchiola/deepgaze Follow instructions as shown in github to install the prerequisites

Install branch 2.0

Replace face_detection.py in /deepgaze/deepgaze with version in this repo

Install dlib for python

Download dlib tar from http://dlib.net/ and follow instructions from http://dlib.net/compile.html for using dlib in python

Troubleshooting

In the case of an error with cv2 check, opencv library imported in python, on terminal

by running

python
import cv2
cv2.__version__

there is the chance that ros kinetic will install cv 3.3 over the 2.4 required by the algorithm to perform properly

Installing libraries on Python

When installing a lib with pip make use of the

sudo -H pip install ... etc