Complete Home Assistant Voice configuration guide

1,644

61 0

Published 2024-03-25

Unlock the full potential of your smart home with voice control! In this comprehensive tutorial, I dive into setting up voice assistants using ESP32 S3 Box 3 satellite devices with firmware crafted in ESPHome. Starting with English, we guide you through configuring these satellite devices to manage my Home - from lights, and roller shutters, monitoring who's home, checking the weather, and even adding items to my shopping list - all through voice commands. Venturing beyond the basics, I explore the challenges and solutions of implementing voice control in other languages, addressing unique hurdles such as voice recognition, language declension, and more - demonstrating that in my Czech native language (with the subtitles where I give the commands in Czech).
Whether you're looking to enhance your home automation system in English or Czech, this video offers innovative solutions to common and language-specific challenges. Dive in to make your home smarter and more responsive!

00:00 - intro (x2)
01:33 - installing voice asistabnt to ESP32 S3 Box 3
05:15 - configure Assist Pipeline (EN, Cloud)
06:44 - device configuration
07:51 - testing cloud (EN)
09:43 - configuring local add-ons (EN)
11:45 - testing local add-ons (EN)
13:31 - testing cloud (CZ)
16:38 - adding aliases for declension
18:14 - testing cloud with added aliases
19:22 - local add-ons (CZ)
21:27 - outro

All Comments (6)

@tazanelli 1 month ago

Thanks to you I am now ready to get rid of my google assistants and leave without fear of being listened to. And I have also learnt that Czech would be a hard language to learn!
@ppirky 1 month ago

Great intro, well done!
@harimax113 yesterday

hello I saw your video but mine when I flash it does not arrive at the wifi request need help thank you
@supperka2 1 month ago

Ahoj, Václave, právě řeším to samé. V českém jazyce rozumí HA až když nastaviš model ve whisper na médium, ale potom je tam prodleva až 10-30s záleží na HW na kterém probíhá rozpoznání hlasů. Pokud přijdeš na možnost jak to vylepšit určitě dej vědět. Jinak super práce a pěkné video. 👍
@weitanglau162 1 month ago

are you able to train your own model for on device wake word using microwakeword?