This job ad has been posted over 90 days ago...
Hidden Markov Model
at
Pre NeuroCyberProsthetics
(Anywhere)
I would like to hire for some help on aplication of open source software.
So good documentation on implementing this so it can be expanded for more states or more observations.
Also this software must be tested by example code or example inputs..
Hi I seen GNU GPL Hidden Markov Model software examples that can be compiled for linux I am useing the maple r5 a ARM cortex m3 .
It has a arduino like IDE or gcc c comand line compile option.
http://www.kanungo.com/software/software.html#umdhmm
It would be interesting if this could process 500 samples per second on that 72 mhz 32 bit processor there is a more memory version of it that is in the works.
The sensors I am working on do have the possibility to shovel the data up a high speed usb port so processing on a i7 or such laptop or desktop running linux with only the terminal is ok.
The data: there 32 bits of information 24 bits samples.
I would like this for a GNU GPL open source project I am working on as a hobby project.
But I am willing to pay to get this part of the project documented and made easy.
I read in "Continuous Myoelectric Control for Powered Prostheses Using Hidden Markov Models"
I have read papers on how the HMM was used to find patterns in bioelectric signals and control prosthetics in realtime based on sensor observations and hidden states the model shows the probability of it being in.
I would like to know how long it would take and how much and if it is not done by then you hold onto what you did and we renegoceate.
So good documentation on implementing this so it can be expanded for more states or more observations.
Also this software must be tested by example code or example inputs..
Hi I seen GNU GPL Hidden Markov Model software examples that can be compiled for linux I am useing the maple r5 a ARM cortex m3 .
It has a arduino like IDE or gcc c comand line compile option.
http://www.kanungo.com/software/software.html#umdhmm
It would be interesting if this could process 500 samples per second on that 72 mhz 32 bit processor there is a more memory version of it that is in the works.
The sensors I am working on do have the possibility to shovel the data up a high speed usb port so processing on a i7 or such laptop or desktop running linux with only the terminal is ok.
The data: there 32 bits of information 24 bits samples.
I would like this for a GNU GPL open source project I am working on as a hobby project.
But I am willing to pay to get this part of the project documented and made easy.
I read in "Continuous Myoelectric Control for Powered Prostheses Using Hidden Markov Models"
I have read papers on how the HMM was used to find patterns in bioelectric signals and control prosthetics in realtime based on sensor observations and hidden states the model shows the probability of it being in.
I would like to know how long it would take and how much and if it is not done by then you hold onto what you did and we renegoceate.
1
person has replied to this posting.
You must be logged in to reply to postings.
Published at 14-01-2011
Viewed: 332 times
Viewed: 332 times
FEED