Skip to content

Overview

Cyborgscode edited this page Jan 18, 2024 · 3 revisions

Welcome to the Personal-Voice-Assistent wiki!

You see the source repository and wonder, what can this assistant do for you, here are some links to videos taken in the past:

Demonstrations

Basic voice control demo End of 2021:

https://static.bloggt-in-braunschweig.de/PVA-Dezember-2021.mp4

PVA is using ChatGPT to fill it's gaps, which means, if it's not meant directly as a command, and no local configured reply, it's send to ChatGPT:

https://static.bloggt-in-braunschweig.de/PVA%20-%20ChatGPT.mp4

PVA controls Netflix WebApp:

https://static.bloggt-in-braunschweig.de/PVA%20-%20Netflix%20Control%20-%202k.mp4

PVA controls a cluster of different Linux devices and send Videos from the main desktop pc to different clients. You see a videoconference on the right, with the local pc and 2 clients, just for observation when and what happens on the clients desktops.

https://static.bloggt-in-braunschweig.de/PVA%20-%20Streaming.mp4

Clustercontrol is automatic, for all configured clients. They need different bitrates for video streaming and have different screen sizes. If a client is available it is auto-detected. All cluster clients combine their microphones and speakers to a virtual device, so if you tell pva to play something, you can direct it on the virtual pulseaudio device and it's play on all devices. That way you can wander around in flat and still give commands to you PVA.

Basics

basic voice functions are:

  • Play music / control the player
  • View videos / control the player
  • Search for files and open them
  • Weatherservice
  • Reminders
  • Start and Stop Apps
  • dictate emails which are delivered via Thunderbird.
  • Take telephone calls via SIP via Twinkle
  • Call people via SIP via Twinkle
  • observe and read emails in IMAP mailbox(es)
  • CalDav and CardDav support
  • Websearch

186 configured commands atm. ( 18.1.2024 )

Text-to-Speech Systems

  • Espeak
  • Mbrola
  • Pico2Wav
  • MaryTTS
  • GTTS
  • Pieper
  • SamsungTTS (additional app needed)

Speech-to-Text Software

Alpha Cepei's VOSK ( see readme ) 18+ different languages needs python

but you can use any other system if you like, as long as you can send the "text" to PVAs network service, which is quite easy.

Supported Systems & Devices

Desktop:

Fedora Linux and all RPM utilizing OSs via own REPO Any Linux with JAVA & PulseAudio & OpenSSL Windows, if you are willing to rework the config for different tool names and options. MacOS, if you willing to rework the config for different tool names and options.

Mobile:

Pinephone with LINUX

Toolkit

OpenJDK JAVA & Python (required) Alpha Cepei's VOSK (required/replaceable) OpenSSL ( replaceable ) OpenSSH ( optional ) PulseAudio / Pipewire ( required ) FFMPEG ( optional for streaming )