Build your own Google Assistant with Raspberry Pi 3

Build your own Google Assistant with Raspberry Pi 3

  

As voice recognition and natural language analysis have improved, the availability of digital personal assistants, from Bixby to Alexa or Cortana to Google Assistant, all leading technology companies continue to develop digital assistants. Recently, Google released its Wizard API for Raspberry Pi 3.
Google and the Raspberry Pi Foundation launched the "Voice Kit" with a special piece of hardware called Voice HAT (basically a speaker driver and microphone), plus great software to build just that, a Google Home Clone.However, getting the Voice Kit has been impossible. It was supplied on the cover of MagPi, the official magazine of Raspberry Pi, and was sold in hours. 

What do you need:

To build your own digital assistant you will need:

Raspberry Pi 3 http://amzn.to/2skfFO5
A 16GB microSD card http://amzn.to/2rwLjty
A speaker with a 3.5 mm connector http://amzn.to/2qMmztn or http://amzn.to/2rwgipJ
A USB microphone http://amzn.to/2skhUAL
A simple switch http://amzn.to/2qGBGp5
A Box http://amzn.to/2sk2D3b
Top-cofrLD Ultra Small PAM8403 5V http://amzn.to/2qGMzY4
3.5mm Male to Male Stereo Audio Cable http://amzn.to/2qGiNTe
Wires Jumper Cable http://amzn.to/2qGAnqf

Let's do is take the Voice Kit software and modify it to work with the internal sound card of the Pi and a USB microphone. Then we will make a switch (to activate voice input) and put everything in a box.The first thing to do is download the microSD card from the Voice Kit for the Raspberry Pi. You can download it directly from the Google AIY project site. Once the .img.xz file has been downloaded, you must write it to the microSD card using a card-writing utility. The Etcher.io tool is a good choice for this, as it is available for Windows, MacOS and Linux. You do not need to unzip the image file, Etcher will take care of it for you.Google AIY project site.  Etcher.io  Recommended Programs:

1. SSH-Client
2. VNC

Create a project

Open an Internet browser and navigate to the cloud console.

GCP uses projects to organize things. Create one for your voice recognition box.In the Cloud console, click the drop-down button to the right of the "Google Cloud Platform"From the drop-down menu, click Create ProjectEnter a project name and click CreateOnce the project has been created, make sure that the drop-down menu has the new project name
1.2. Enable the Google Assistant API
 In the cloud console, enable the "Google Assistant API".In the cloud console, create an OAuth 2.0 client In the API Manager> CredentialsClick on Create Credentials and select OAuth Client ID

• If this is the first time you create a customer ID, you must configure your consent screen by clicking the Configure consent form. You must name your application (this name will appear in the authorization step)

Select Other, type a name to help you remember your credentials, and then click Create.A dialog box will appear. Click OK. In the Credentials list, find your new credentials and click on the download icon (Download icon) on the right.Note: If you do not see the download icon, try to widen the browser window or zoom out.Find the JSON file you just downloaded (client_secrets_XXXX.json) and rename it to assistant.json. Then move it to /home/pi/assistant.json 

Start dev terminal:

 Double-click the "Start dev terminal" icon on the desktop. In the terminal window, type sudo leafpad/boot/config.txt and delete # in front of the line dtparam = audio = on and insert a # in front of the two lines below it.  exit nano.From the previous step, the last lines of /boot/config.txt should look like this:

# Enable audio (loads snd_bcm2835)

dtparam=audio=on

#dtoverlay=i2s-mmap

#dtoverlay=googlevoicehat-soundcard

Audio

The next step is to get the audio to work. You will need a speaker with a 3.5 mm jack and a USB microphone. I did not have a dedicated USB microphone on hand so I plugged in a spare webcam that I had and the Pi was able to use the webcam microphone as a standalone microphone! If you need a USB microphone, try it. Connect the speaker to the 3.5mm jack on the board and connect the USB microphone to one of the USB ports.Start dev terminal :

sudo nano /etc/asound.conf



Delete all content and change it for this:



pcm.!default {

  type asym

  capture.pcm "mic"

  playback.pcm "speaker"

}

pcm.mic {

  type plug

  slave {

    pcm "hw:1,0"

  }

}

pcm.speaker {

  type plug

  slave {

    pcm "hw:0,0"

  }

}

Start dev terminal:

nano /home/pi/voice-recognizer-raspi/checkpoints/check_audio.py



And let's change this line VOICEHAT_ID = 'googlevoicehat'

THEREFORE: VOICEHAT_ID = 'bcm2835'



In case it does not work we will have to do tests following these Configure and Test the Audi

The switch

The circuit for the switch is as follows: Once you have built your circuit, you must connect one side of the switch to GPIO 23 and the other side to ground, which is conveniently located on the next pin to the left of GPIO 23.If you can not build the switch, even just short circuit GPIO 23 and earth will have the same effect!

LED:

LED negative to GPIO 25 and positive to GPIO + 5v

Try it:

You should now have everything you need to try Google Assistant on your Raspberry Pi. All you need to do is run the following command from the dev terminal:

src/main.py 

To start automatically:

Edit rc.local
Command from the dev terminal:

sudo nano /etc/rc.localsystemctl enable voice-recognizerCntl-X, Return to save.

 

more information Voice Kit