Hacking the Google AIY Voice Kit - Part 1

This is from way back but I discovered a version of the AiY kit on my shelf unopened (got it from Microcenter here in Pennsylvania a long time ago). I want to do speech recognition for my DJI Tello drone for which I created an ESP32 controller. I now have some PCB prototypes of the controller and I can use a serial connection to the controller to send commands using pyserial.

By any chance is the source code for the Part 1 and Part 2 videos and articles available. The dbot.ws site seems defunct.

Also, can the Pi 4 be used with the AiY hat board?

@jjs357 

Not sure if this is what you want but take a look at ...

part 1 - https://dronebotworkshop.com/hacking-google-aiy-voice-kit-part-1/

part 2 - https://dronebotworkshop.com/hacking-google-aiy-voice-kit-2/

@will Thanks. I do have those links. I was under the impression (maybe wrong) that code in support of using the AiY kit was also available. These days Bill provides a url to a zip file of code resources for most of his videos. Maybe i am mistaken.

@jjs357 

I didn't follow that video because I don't do Python, but it seemed that most of the code was provided by third parties and involved much downloading, fixing, updating and re-downloading. You needed a kit for the hardware and lots of patience for the software.

The hacking appears to be in part 2 and Bill's code is included. I doubt if he's legally able to provide all of the other source code, you'll have to set up the background system by following the instructions he gives.

Sorry, there's probably no clear, easy path.

@will OK, thanks for the update. I will at least attempt to make something of the kit. I have a fair bit of Python experience to build on. The last Raspbian special Pi image from Google was released in April of 2021. So we are almost 2 years removed from that time. The version 2 AiY kits are still for sale though and these include a Pi Zero in the box. I have an older V1 kit.

@jjs357 It's all there. Look under Resources or search.

@zander Thanks. This time I scrolled all the way to the bottom of the Part 2 article and there I see the code Zip. I was learning about this library earlier today:

https://github.com/Uberi/speech_recognition

And the Google Cloud approach was mentioned there. The older Google Web API for speech is still somewhat supported and I got some little examples to work with it. But the Google cloud demo is pretty impressive -- near real time recognition of spoken words.

@jjs357 No need to scroll to the bottom, there is a Table Of Contents on the right side, just scroll down until you see Resources and click that link. It's all there, but be aware Bill does not often go back and update old resources if the underlying code has changed or is obsolete. That is my somewhat clever way of saying if it doesn't work don't expect any help here, you need to go to the original authors.

@zander Yes, the ToC item does also show resources. I guess I had the ToC view off. The Google Assistant API is being End-of-Life'd in June 2023 anyway, so I will learn as much about speech recognition as I can in the mean time.

@jjs357 Where is the setting to turn Off or On the TOC view?

@zander I see this:

 So I guess I mean hidden and not off.

@jjs357 Are you punking the old man? No such thing on my page.

@jjs357 I never noticed that 'mini' TOC, I am talking about the always there TOC on the right a little way down.

@zander Yes, that's there too! I guess I scrolled by too quickly -- the ToC remains close to the top right below the FAQ graphic. In any case, I now know multiple places to look for more information. Appreciate the advice.