3 Talking-Pi Personalization Manual
3 Talking-Pi Index 1. Introduction 2. Setting up the Cloud Speech API 3. Custom Wake Words 4. Custom Commands 5. Support Ausgabe 27.12.
Talking-Pi 1. Introduction Dear customer, Thank you for choosing our product. The own programming and modification of the system is an essential and important part of this project. In the following steps, we will explain step by step how to set up the Google Assistant API, make your Talking-Pi ready for use, and how to program and modify your Talking-Pi according to your own needs. If, however, problems should occur while working with your Talking-Pi, please feel free to contact us.
3 Talking-Pi 2. Setting up the Cloud Speech API With our pre-installed Talking-Pi operating system, the Talking-Pi waits patiently to be activated. You can download the image here. Install the image on a micro SD card (a capacity of at least 8GB is recommended) and then insert it into your Raspberry Pi. With a modified Cloud Speech API, 80 languages are available. You can also enter your own wake words and commands in the system.
3 Talking-Pi Next, select the option "Billing" from the left menu. Add a new payment method and follow the instructions. After adding a new billing option, return to the billing menu and make sure your newly created project is linked to this payment option. To link your project to a payment option, you can simply select the three items. Next, the required services are activated. Use the icon marked in the picture to add a new project to your list.
3 Ausgabe 27.12.
3 Talking-Pi In the API library that now opens, search for the term "Google Cloud Speech API". Select the found API and activate it. Now select the item "Access data" in the "API & Services" overview. Ausgabe 27.12.
3 Talking-Pi Select "Create login data" and create a "Service account key". Here you define the project and viewer as roles: Select "JSON" as the key type. Ausgabe 27.12.
3 Talking-Pi Now create your access data by clicking on "Create". The download of your data starts automatically. Rename the downloaded file to cloud_speech. json and copy it to the user directory of your TalkingPis. After a restart, the Cloud Speech API is ready for use on your TalkingPi. Ausgabe 27.12.
3 Talking-Pi 3. Custom Wake-Words The activation can be done either by pressing the button or by a so-called wake-word, which has been integrated into the system before. In the system we have prepared, the talking pi reacts to the two Wake-Words "Talking-Pi" and "Alexa". If these two wake-words are not enough for you, or if you want to personalize your Talking-Pi, you can of course also teach your talking-pi your own wake-words.
3 Talking-Pi You record the wake word three times to adapt it to your voice. Afterwards, you can download the .pmdl file that matches your wake word. Then copy this file to the following folder of your Talking-Pi: /home/pi/GassistPi/src/resources/ In order to activate the wake-word in your talking-pi it is necessary to modify the configuration file of the Gassist Pi library. To do this, type the following command into a terminal on your Talking-Pi: sudo nano /home/pi/GassistPi/src/talkingPi_snowboy.
3 Talking-Pi As in the following example, modify the models array to add or remove your own wake word configuration files. models = ['/home/pi/GassistPi/src/resources/alexa.umdl', '/home/pi/GassistPi/ src/resources/Talking_Pi.pmdl', 'home/pi/GassistPi/src/resources/ [mywakeword].pmdl'] Please note that the phrase "[mywakeword]" is a placeholder to be replaced by the filename of your own wake word.
3 Talking-Pi 4. Custom Commands To add your own voice commands, which the Talking-Pi recognizes as a separate command, it is necessary to first define the command as an expected input. This increases the detection rate. Open the file "googlecloudtalkingpi.py" with the following command: sudo nano /home/pi/voice-recognizer-raspi/src/googlecloudtalkingpi.py Add the following recognizer. expect_phrase command to the existing phrases: recognizer.
3 Talking-Pi elif '[Ihr eigenes Sprachkommando]' in text: subprocess.call(["python","/home/pi/I2C-LCD/EigenerBefehl.py"]) subprocess.call(["bash","/home/pi/TalkingPi_Befehle/Receiver/ EigenerBefehl"]) subprocess.Popen(["aplay", "/home/pi/GassistPi/src/resources/dong.wav"], stdin=subprocess.PIPE, stdout=subprocess.PIPE, stderr=subprocess.PIPE) time.sleep(2) Please be sure to replace the placeholder with your own command. You may also need to correct the indentations after copying.
3 Talking-Pi Please note that the files to be executed must still be created by you in the corresponding folders. You can of course define as many files as you need. In our standard configuration we have defined the following files for execution. These can be used for your personal configuration. Display-Control: File location: /home/pi/I2C-LCD/[YourFile].py Example: import lcddriver from time import * lcd = lcddriver.lcd() lcd.lcd_clear() lcd.lcd_display_string("+++ Ausgabe +++", 1) lcd.
3 Talking-Pi 5. Support We also support you after your purchase. If there are any questions left or if you encounter any problems, feel free to contact us by mail, phone or by our ticket-supportsystem on our website. E-Mail: service@joy-it.net Ticket-System: http://support.joy-it.net Phone: +49 (0)2845 98469 – 66 (11- 18 Uhr) Please visit our website for more informations: www.joy-it.net Ausgabe 27.12.
