[microsound] image sonific...

[Dara Shayda]

(The thread got too long and got a moderator message so I cut it to a smaller piece)

Hello Balzas

There are two sources of video and images for the phonemes:

1. The usual facial gestures as you are doing your self. But I think you need to wear some make-up i.e. some dots on your face which you track in 2D or 3D depending on how sophisticated you want to get, see this for ideas:
http://www-prima.inrialpes.fr/FGnet/data/01-TalkingFace/talking_face.html  you can make a simple system for yourself by painting your face with highly contrasting dots and then record the video and extract their 2D or 3D motion. Lots of papers on that on the net. 

2. http://www.youtube.com/watch?v=Q-PZlj0UH0c  you can borrow from these videos, easily we can get sequential frames and then image synth them. You also have the actual sound from the larynx which can overlay the image synth sounds

I will write you more offline

D


On 2012-01-10, at 2:03 AM, Kovács Balázs wrote:

> Hi Dara,
> 
> sorry for the late response, and many thanks for the support!
> 
> i plan to begin the interaction-based works in the last part of the project, probably
> the next week? i built a basic eye/mouth follower (http://vimeo.com/31281227),
> but i don't have experiences with phoneme recognition... anyway it would be very
> good to include it in the series, so if You have idea, then welcome and i would join
> in! maybe i could also develop a system for the end of the next weak, but as i see
> you have experiences in the field with this wonderful "Beginning" work!
> 
> All the best,
> Balázs
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://or8.net/pipermail/microsound/attachments/20120110/6cb576ee/attachment-0001.html>