stereo

I was wondering if anyone out there managed to do a handwriting recognition.

- User writes on trackpad/touchscreen 1 letter per time

I was reading some older posts (without getting any good tips) and the article of Paul Kolesnik and Marcelo Wanderley "Implementation of the Discrete Hidden Markov Model in Max / MSP Environment" in which they use the dishmm object for the training, optimal sequence of states and recognition. (That object I cannot find online anymore)

Hello Everyone,
I was wondering if anyone out there managed to do a handwriting recognition.
What I need to do:
- User writes on trackpad/touchscreen 1 letter per time
- recognize letter
- then I trigger a specific sound.

I was reading some older posts (without getting any good tips) and the article of Paul Kolesnik and Marcelo Wanderley "Implementation of the Discrete Hidden Markov Model in Max / MSP Environment" in which they use the dishmm object for the training, optimal sequence of states and recognition. (That object I cannot find online anymore)
Any help ? or hints where to start?
Thank you in advance


handwriting-recognition

in the FTM Library there is a patch that might do part of what you are after.

it's called gesture follower. you write a shape on a graph while recording the audio of each movement and than by repeating on the shape it will play back the sound in order of each movement.

check this youtube video it explains better than i do:

https://www.youtube.com/watch?v=2dSU8MeJlz4

this is where you download the FTM Library, which includes the gesture follower.

in the FTM Library there is a patch that might do part of what you are after.
it's called gesture follower. you write a shape on a graph while recording the audio of each movement and than by repeating on the shape it will play back the sound in order of each movement.
check this youtube video it explains better than i do:
https://www.youtube.com/watch?v=2dSU8MeJlz4

this is where you download the FTM Library, which includes the gesture follower.
http://ftm.ircam.fr/index.php/Download

Thanks for the hint! I will try to have a look at the part of "recording" the gesture and "matching" up features..

Funny the sound part :); even though I am just interested in triggering specific sound feedback when the letter is recognize!

Thanks for the hint! I will try to have a look at the part of "recording" the gesture and "matching" up features..
Funny the sound part :); even though I am just interested in triggering specific sound feedback when the letter is recognize! 

The library is cool, but the object I really need *mnm.follower* does not work for me.

"mnm.follower: unable to load object bundle executable"

I am using Max 7 (open in 32bit) and I tried also with Max 6.1 (32bit as well) and the library -> FTM & Co for Max 6.1 and Max 7 32 bits. It is not a problem of path because the other external are seen and works fine :(

The library is cool, but the object I really need *mnm.follower* does not work for me.
Error I get:

I am using Max 7 (open in 32bit) and I tried also with Max 6.1 (32bit as well) and the library -> FTM & Co for Max 6.1 and Max 7 32 bits. It is not a problem of path because the other external are seen and works fine :(
Any suggestions?
Thanks
