Asked — Edited

Respeaker Raspberry Sound Location Question

User-inserted image

I have been very interested in having my robot look in the direction of a person speaking. In the above picture you will find a far field Realspeaker mounted on a Raspberry Pi.  The system will allow you to output sound location so the robot head can follow the sound of the person speaking.  Since Ez-Robot microprocessor can now work with the Raspberry Pi, I thought this would provide state of the art sound location but would require loading all the software to accomplish it. My question is 1. Does anyone know how to explain to a novice how hard would it be to install and edit the required software to make this work and 2.  Will this work with Ez-Robot microprocessor.  I believe it will.  It will cost around $65.00 in equipment.


ARC Pro

Upgrade to ARC Pro

Don't limit your robot's potential – subscribe to ARC Pro and transform it into a dynamic, intelligent machine.

#2  

Are you sure this will be working...dont you think a camera is the better option? I think its a good idea to have the robot looking to the speakers direction, I would really like this feature...but have you seen any working Implementation?

#3  

I think that after getting this work maybe there would be a way of combining video and sound.

PRO
USA
#4  

@Ellis, Are you using the ARC Designer Pi Beta ? Which Raspberry PI version do you have  ?

#5  

Raspberry Pi Zero, Raspberry Pi 1 B+, Raspberry Pi 2 B, Raspberry Pi 3 B and Raspberry Pi 3 B+ Are all comparable with the respeaker.   I want to use the one in the picture. I think this Pi and respeaker will also bring a lot of capabilities beyond sound location. The link on the respeaker is above.

PRO
USA
#6  

@Ellis, I read the product page, the target audience is Raspberry PI hardware running the ARC Designer Pi Beta on the Raspberry PI.

Can you share your experience with Raspberry PI and ARC Designer Pi Beta ?

#7  

I would love to but I have very little experience with both.  I was checking if it would work and if I would be able to do it with the little experience I presently have.  I know it seems to be very documented in Sythiam Designer Pi Beta and the Pi software but all the downloading and linking software caused me to have  caution.

#9  

Its kind of funny because they seem to use the same hotword detection for their microphone array...:D

PRO
USA
#10  

The product is interesting if you are building an Alexa clone, to work as expected you need to lay flat horizontal, for a robot you need to put on top of the head.

Another mic array https://www.seeedstudio.com/ReSpeaker-4-Mic-Linear-Array-Kit-for-Raspberry-Pi-p-3066.html you can attach vertically to the robot.

Sound localization is not a new thing, Tony Ellis did that before: https://synthiam.com/Robot/Introducing-Altair-Ez-2-Robot-563

There are some linux open source scripts for microphone arrays, and there is a ROS solution:  https://github.com/uts-magic-lab/hark_sound_localization

You can write some python scripts and integrate with ARC.

Installing, managing and maintaining these products requires some Linux and raspberry PI knowledge.

#11  

Thanks ptp.  I am going to review this information to see what I need to do.