[Larry]

My name is Larry, and I am a retired auto mechanic from Canada. I seen a video on utube about a humanoid head with A.I. and thought I might try to make one myself. I have managed to make the head frame and and have started to mount servos in it.
I have a chatbot from pandorabots and would like it to control the eyes and mouth. I have experimented with sound control to a arduino to move the jaw, but I have to figure out speech recognition for the pandorabot. Wish me Luck!

[infurl]

That sounds like a great project, I hope we'll get to see some pictures or video soon. I've had a look at speech recognition for my own robotics projects and I've got a friend here in town who has put in considerable effort trying to get speech recognition to work with open source software on a Raspberry Pi, but without a lot of luck so far.

However, I've seen videos on youtube and project write-ups that suggest it is pretty easy to do speech recognition with something like a Raspberry Pi if you install the Android operating system on it and hook it up to the cloud, e.g. something like Amazon's Alexa service. Speech recognition seems to be something that just about any old smart phone can do superbly nowadays, if you don't mind being connected to the internet for it to work.

[Larry]

I thought I would use a windows tablet to connect to the internet, and use a sound output from the earphone jack to a LM383 amp to boost the signal, run the right channel to the arduion to control the jaw servo, and the left channel to a speaker. If the chatbot talks, the jaw should move in sync.
How do you attach images to a post, I would like to show you how I made this thing out of scrap metal and car body filler (Bondo), I looks pretty good, I thought.

[Korrelan]

Hi Larry and welcome 

You could cut the Arduion out of the equation altogether by using an optoisolator driven straight from the audio channel. A simple darlington amp or even single transistor gate could then trigger a relay or servo directly. You could easily wire a DPDT relay to reverse the polarity and return the jaw servo back to a closed position (microswitch closed sensor).

For adding images to the forum I personally use Imgur. Set up an account and upload your pictures, they provide a link for boards etc… just copy and paste.

Good luck with the project.

 

[Art]

Hi Larry and welcome aboard!

It's always nice to receive new visitors and hear / see of their projects. Best of luck! Keep us posted on your progress.
You might also wish to check out the work at: https://inmoov.fr/

Shown is it's head. (yes, 3D printed - as is the whole robot).

[Larry]

I uploaded some pics to Imgur and here is a link, Let me know If it works.
It is a bad link, Sorry

[Freddy]

HI Larry and welcome

I'm getting "LarryLea's images are not publicly available.".

Have you looked at the Chromium Project speech recognition ? I'm using it with Unity and it works quite well.

[Larry]

http://imgur.com/pBt8mve
http://imgur.com/aIuYAPt
http://imgur.com/2VfPyRa
http://imgur.com/q5bkVHf
http://imgur.com/ymQ6IJ1
http://imgur.com/G59BESF
Try these image links and let me know if it works,(I found a button to generate image links)
And I would love a Imoov, but I would need a 3D printer.
I will do a search  for chromium project, Thanks

[Korrelan]

Yup the links worked.

Project is looking cool.

If you need your image to display on the topic click on the uploaded image in imgur and copy this link into your message text. (marked in red).



 

[Larry]

Thanks Korrelan