voice recognition

pantera · 07-01-2004, 12:42 AM

can someone please tell me how to do programming in linux to do voice recogition only.i just want the program to take my voice store it somewhere and then when i speak again it will check my voice and disppay error if it doesnt match

pycoucou · 07-02-2004, 07:48 AM

I would say it is a kind of specialised questions. Have you check on the net the research center in this field? They may provide some demos you could use... From scratch, it's going to be a huge work!

Ashkhan · 07-03-2004, 05:21 AM

Speech recognition is very complicated. Complex programs use Dynamic Time Warping or Hidden Markov Models for recognition of words or continuous speech (speaker independent). It should be easier to do this for 1 speaker and one word commands but it is still very difficult task. This includes creating the database of words (their parameters based on LPCC analysis), training of your algorithm,...

as pycoucou says try to google something up.