Ep04: Voice Assistant using Quarky Intellio | ESP32-S3 with Mic and Speaker | IoT Projects
Watch on YouTubeVideo summary
This video introduces a comprehensive guide to building a voice assistant using the Quarky Intellio device, which features an ESP32-S3 microcontroller equipped with an omnidirectional MEMS microphone and a high-resolution 1W speaker. The tutorial begins by explaining how to power on the device and establish a connection via Wi-Fi or Bluetooth, noting that users must ensure their network operates on the 2.4 GHz band for compatibility. Once connected, the project demonstrates fundamental voice control capabilities by programming the device to record audio when a specific button is pressed and play it back upon another press, effectively turning the hardware into a simple voice recorder and player.
The demonstration then progresses to more advanced functionalities, such as playing music files stored on an SD card using number buttons for navigation. The creator further enhances the user experience by mapping these buttons to control playback directly with the L and R keys, allowing users to start songs or stop them without needing to speak. A significant leap in interactivity is achieved by integrating speech recognition extensions, which enable the device to understand natural voice commands like "play song one" or "stop," eliminating the need for physical button presses for basic media control.
Beyond music playback, the video showcases an interactive desk companion project that utilizes speech recognition to engage in basic conversations with pre-determined responses, such as answering questions about mood or favorite foods. To make this interaction more engaging, a TFT display is added to show animated emotional reactions corresponding to the dialogue. The tutorial concludes by emphasizing the core principle of these projects—listen, decide, and respond—and encourages viewers to expand their skills toward creating smart home IoT devices that control lights or fans via voice, while also teasing future content that will combine cameras with audio features for even smarter AI applications.
Read the full video transcript
Hey, it's a nice day. Can I play your
favorite song?
>> Oh, yes. Play number five.
[bell and music]
Okay, Intellio, stop.
Hey. Hey everyone, this is Ai and
welcome back to our quirky Intelio
series. Today we are exploring how Intel
can listen, talk, play your favorite
music and interact with voice commands
using its microphone and speaker. So
let's get started. Intellio is equipped
with omnidirectional [music] MEMS
microphone and a powerful highresolution
1W speaker
now. Step one, let's power on Intelio
[music] and connect to Pictoblocks via
Wi-Fi router method. If you are already
connected via pictoblocks, so you can
totally skip this step. Open pictoblocks
and choose the block holding
environment. Then click on board and
select Intelio as a board and choose the
first Wi-Fi router option.
Now enter the same Wi-Fi name and
password to which your laptop is
connected currently and make sure your
Wi-Fi is running on 2.4 GHz band as
Intel supports only 2.4 GHz network.
[music]
To know more about this connection
method, you can obviously watch our
previous video where we have explained
it in more detail. Check if your systems
Bluetooth is turned on or not. If it is
turned on, great. [music]
Now, as the next step, click on next. Do
select your Intel serial number which is
next to the power switch over here.
[music] Mine is AD2A.
Intel's RGB LED will turn green as you
can see on Bluetooth connecting
successfully. [music] Now it RGB LED
will blink yellow while connecting. Once
the connection is successful, you will
see the RGB LED glow yellow
continuously. [music]
Then click on go to editor. So let's
start with something super simple.
[music] I want that when I press the
left button, it starts recording my
voice for let's say 3 seconds. And when
I press the R button, it stops recording
and it plays my recording. Okay, so
let's code [music] for that. So first
I'm going to put when green flag click
as an event and then [music] I'm putting
a forever.
Inside this we will put two if blocks.
First if block is for when [music] L
pressed and the second if block for R
pressed.
Now for [music] the first if block I'm
going to go to mic and I will take this
block [music] to start recording in a
flash memory.
Then [music] I wanted to play. So for
that I will go to speaker and I will
take this block to play from the flash
which means the current memory. [music]
Now let's try it. I'm going to press
green flag.
And now I will press the L button.
I love Intel.
Now let's press [music] R button.
>> I love Intel.
>> Yes, it worked right. Okay. Now let's
level up by playing the songs using an
SD card. So I have added SD card in a
card reader and I have added this five
[music] songs in this. All right. Now I
want to play them in pictoblocks. Let's
see how we can do that in coding.
[music] So as you can see the code over
here. When I press the one button, it
will play the first song. When I press
number two, it will play the second
song. [music] When I press number three,
it will play the third song. And so on.
So let me just add the SD card
in Intellia over here.
Let's see if this works. I'll [music]
press number one.
>> [music]
>> Okay. Now to stop it, I'll press space.
Now let's play another song. I'm
pressing number three.
Old
[music]
>> space.
>> That was interesting. Right now, we can
also level up by mapping the buttons L
[music] and R. So, if I press L button,
it will play the first song and when I
press R button, [music] it will stop.
Let's do that.
I'm changing this event to when L button
pressed it will place first song.
[music]
When R button pressed, it will stop the
song.
Now when I press L button, let's see.
[music]
I'll press R button. Now
it stopped. [music] Interesting. Now we
can also level up by adding voice
recognition in this. For example,
[music]
if I say play song one, it will play the
first song. If I say number two, it will
play the second song. And [music] when I
say stop, it will automatically stop. So
for that, we will also have to add a
speech recognition extension in this.
Now let's check out the code over here.
As you can see, I have added speech
recognition extension [music] and these
are the speech recognition extension
blocks. Below that, in the forever loop,
I have added some if conditions. When I
say one, it will play song number one.
When I say number two, it will play song
number two and so on till five. When I
say stop, it will stop playing the song.
So, let's try and check this. I'll press
the green flag
one.
Now I'll say stop.
[music]
Interesting, isn't it? So this was your
voice recognized music player.
So I have also created this interactive
desk companion where we are using speech
recognition for some [music] basic
conversation and which it will reply
from preddecided responses through
speaker. So let's see if it works. I'll
press the green flag. [music]
How are you?
>> I am superb.
>> Great. How was your day? [music]
>> It was fabulous.
>> What is your favorite food? [music]
I love battery.
>> Wow, that was quite interesting, right?
[music]
And oh, I have also added this TFT
display with some animation of emotions
to make it more interactive. And this
was an advanced project where Intel
listens, understands the command and
give controlled responses. Now, you can
also make something like IoT house with
Intel in which it turns on the fan or
light based on your voice command. Fan
off.
All these projects follow the same idea.
Listen, decide, and respond. If you can
do the activities from today's video,
you're already on your path to create
more innovative voice-based AI projects
where you wouldn't need pre-recorded
responses anymore. [music] We can try
creating something like Alexa.
As an activity for you, you can try
making walkie-talkie using picto blocks,
but do let me know in the comments below
your project names. In the next video,
we will combine camera, mic, and speaker
to create [music] smart AI desk
companion. Spoiler alert, we will also
use chat GPD extension in that. [music]
So, make sure you like, subscribe, and
hit the bell icon. Until then, happy
learning.