UBUVoiceAssistant es una aplicación de escritorio que utiliza un asistente de voz, Mycroft, para que, mediante comandos de voz o texto, el usuario pueda interactuar con una plataforma Moodle.
Desarollado por Rodrigo Garcia Martin para el trabajo de fin de grado 2021 de Ingeniería informática en la Universidad de Burgos
Read this in Spanish here, from page 29.
- Windows 10 (recommended to have latest version) or Ubuntu (recommended latest LTS or stable)
- 4GB RAM for Ubuntu, 6GB for Windows
- 5GB free on disk
You need a CPU that supports AVX technology. Check if you have it on Linux: cat /proc/cpuinfo | grep avx
If you are using a virtual machine and you have Hyper-V enabled on Windows, Ubuntu won't detect AVX and you will need to disable Hyper-V using these commands in a Powershell as admin:
bcdedit /set hypervisorlaunchtype off
DISM /Online /Disable-Feature:Microsoft-Hyper-V
Disable-WindowsOptionalFeature -Online -FeatureName Microsoft-Hyper-V-Hypervisor
Enable them back after you have finished using the project using these:
bcdedit /set hypervisorlaunchtype auto
DISM /Online /Enable-Feature:Microsoft-Hyper-V
Enable-WindowsOptionalFeature -Online -FeatureName Microsoft-Hyper-V-Hypervisor
You will need to enable WSL. Follow this guide, steps 1 to 5. On step 6, install Ubuntu LTS or latest from the Microsoft Store.
After that, you need to download and start VcXserv from here. Don't forget to disable access control. Add these lines to ~/.bashrc:
Only if using WSL 1: export DISPLAY=:0
Only if using WSL 2: export DISPLAY=$(awk'/nameserver / {print $2; exit}'/etc/resolv.conf 2>/dev/null):0
export LIBGL_ALWAYS_INDIRECT=1
Download Pulse binaries from here. Edit etc/pulse/default.pa and add these lines:
load-module module-native-protocol-tcp auth-ip-acl=127.0.0.1
load-module module-waveout sink_name=output source_name=input record=1
Add this line to etc/pulse/daemon.conf:
exit-idle-time = -1
Finally, add this to ~/.bashrc:
If using WSL 1: export PULSE_SERVER=tcp:127.0.0.1
If using WSL 2: export PULSE_SERVER=tcp:$(awk '/nameserver / {print $2; exit}' /etc/resolv.conf 2>/dev/null)
Run bin/pulseaudio.exe in a terminal
You will need to do the Ubuntu part too
Get the source code for the last release and extract it anywhere. cd into the extracted folder and run sudo ./install.sh install. After it finishes, you will have an icon in the Ubuntu's app launcher or you can run UBUVoiceAssistant in a terminal.
If it gets stuck in an interactive part, run sudo ./install.sh uninstall and then sudo ./install.sh install --manual
Run these commands:
sudo apt update
sudo apt install git
git clone https://github.com/rogama25/UBUVoiceAssistant.git
cd UBUVoiceAssistant
sudo ./install.sh install
After that, launch the program through Ubuntu's app launcher or running UBUVoiceAssistant
If it gets stuck in an interactive part, run sudo ./install.sh uninstall and then sudo ./install.sh install --manual
If you encounter an error when running the project in WSL that says “ImportError: libQt5Core.so.5: cannot open shared object file: No such file or directory, it's a known bug, and you can run this as a workaround:
sudo strip --remove-section=.note.ABI-tag /usr/lib/x86_64-linux-gnu/libQt5Core.so.5
It's recommended to follow along the usage instructions because it will install automatically all dependencies. Then you will have the UI at /usr/lib/UBUVoiceAssistant, and skills at /opt/mycroft/skills. Mycroft will be located at usr/lib/mycroft-core
It's recommended that you install QtDesigner using sudo apt-get install qttools5-dev-tools and poedit sudo apt install poedit. Also recommended to use mypy and pylint, using the files in the scripts folder.
Detailed and manual way explained in Spanish here, from page 21.