Roboquin Speech Synthesis & Recognition with Alexa & Arduino

Roboquin’s speech synthesis and voice command recognition uses Amazon Alexa. Arduino is used to control its head and arm’s movement.

AdvancedFull instructions provided3,317

Fourth Place

2018 China-US Young Maker Competition

Roboquin Speech Synthesis & Recognition with Alexa & Arduino

Things used in this project

Hardware components

Raspberry Pi 3 Model B

Arduino UNO

LED (generic)

Software apps and online services

IFTTT Amazon Alexa service

Hand tools and fabrication machines

3D Printer (generic)

Story

Human robot social interaction has become one of the active areas of research in robotics. Roboquin is designed and developed as a platform for research in speech synthesis and recognition as well as in several related areas: 1) Internet of Things (IoT) Mannequins, 2) Software controlled speed and pitch as well as the gender and native or foreign accent of mannequin's speech, 3) Interactive robotic mannequin interacting with people in different ways, 4) Social robotics research projects, and 5) Kinematics based movement control in sync with the speech to create a gesture.

Electrical Components and Materials: •Raspberry Pi 3 •Arduino UNO •RPi Camera •Ultrasonic Sensors •Speakers and LEDs •Servo Motors •Tetrix Robotics Kit •Cardboard

Features: 1) The most exciting feature of Roboquin is that RGB LED eyes change color and the Bar Graph LEDs used for lips move in sync with its speech. To achieve speech function eSpeak and Festival software is used, 2) Improvement in the current state of Roboquin’s speech synthesis is planned by implementing voice recognition and voice commands using Amazon Alexa and Google Voice, and 3) Plans are also underway to use the mannequin in an inter-disciplinary course that combines computer technology, social sciences and humanities.

Areas of further research and teaching where interactive mannequin may be used: ●Social robotics research projects ●Display model for wearable & fashion technology ●Mobile device based remote control ●3D physical modeling and design ●Computer controlled system design ●Embedded systems programming

Raspberry Pi Alexa setup instructions:

https://www.raspberrypi.org/blog/amazon-echo-homebrew-version/

https://youtu.be/frH9HaQTFL8

https://github.com/alexa/alexa-avs-sample-app/wiki/Raspberry-Pi

Roboquin Block Diagram

Here microphone and Raspberry Pi/ Alexa Pi are the hardware parts which relates to the WiFi/Internet to Amazon Voice services. Amazon Voice services are using Artificial Intelligence. Amazon Voice services and IFTTT are software services in cloud environment. https://ifttt.com/ is a good website to know more about IFTTT software. It is sending text to Flite/espeak software. Espeak software had some problem, but for the project purpose Flite/espeak both were used. Flite stands for festive lite which convert text to speech and send it to speaker. Speaker is the hardware part and with the help of it Roboquin talks to people. She says that hello my name is Roboquin and more.

Code

Raspberry Pi Linux Commands to Implement Speech Synthesis

Install espeak application using the command: 
$ sudo apt-get install espeak 
Connect the speakers to Raspberry Pi audio output in order to get the speech. Following command is used to test the speech hardware and software: 
$ espeak "This is a test" 
Create the file roboquin1.txt containing the following text: 
hello my name is robo quin.
Use following command to tell Raspberry Pi to say the text that is in the text file:
$ espeak -f roboquin1.txt
The gender, accent, pitch and speed of the voice is controlled by using the following command:
$ espeak -ven-us+f4 -s100 -f roboquin1.txt