NAOsolutions

University Project for enhancing the NAO with python 3.X libraries and python 2.17 with naoqi.

This script in splitted into 2 Folders. The python2 folder should be opened via the dockerfile and the python3 folder provides the possibility to enhance NAO´s features like face recognition and emotion detection from DeepFace.

All Tools used are open-source and can be used for free for further development. Communication between the two enviroments is done via flask-restful API.

setting up everything thats needed

• build docker container with the python.dockerfile and the requirements.txt file and cofigure bind mount to exchange data between docker and the local system
• go to python3/main.py and start the API
• go to python2/main.py and adjust the connection settings (NAOIP & BASE_API). Attention: When using Windows you will need to enter the IP Adress of your WSL (use ipconfig/all for looking up)
• start the script

python2

• actions.py stores all functions related to NAO´s movement (dances, holahoop). To add more movements add a new class.
• dialog.py stores all of the dialogs for the default functionality of this repository.
• functions.py stores all functions related to the main.py scipt (conformation loops, photo capture, audio recording, ...)
• main.py contains all configuration data needed to establish all connections and run the script.

python3

• main.py contains all the functions for creating needed folders and starts the API (Windows: WSL IP in local env)

fileshare

Acts as local fileshare. The python 2.7 enviroment gets access granted through bind mount.

• images contains all the pictures made in one session (gets cleared automatically after every communication)
• knowledge_base contains all saved pictures needed for the face recognition
• names.csv contains all names for the face recognition (empty at default)
• runs.csv contains all data thats gathered through the feature we implemented (empty at default)

Note: There must be exactly the same number of images or entries in the Names.csv and in the knownledge database.

Please feel free to use this Flask and Docker setup for your own features without losing the possibilities provided by the naoqi-library!

Developers

• Böswarth, Lukas
• Dielacher, Marcel
• Kopeinig, Matthias
• Roucka, Alexander