suresh chandra sekaran - 2008-09-22 
         We have trained the system for a vocabulary of 500 words . the training involved speakers in the age group of 19 to 21 and the system works well in recognizing the words spoken by people in the same age group. We have planned to make it speaker independent by adding trained files from people of all age groups from diverse backgrounds .

          Will this alone make the system speaker independent or should i do something more? Do i have any option to implement this in a better way in Sphinx? (I mean changing the config files etc.) .