Getting an error "Input overrun, read calls are too rare(non fatal)" . Mic settings are fine , and i have given the lm and dict on the command line. How do i get rid of this error?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Don't use very large grammar otherwise recognizer can not process incoming speech in time. Probably you also need to optimize recognizer to run faster on your platform. You can play with -beam and -wbeam parameters to make it faster.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I have'nt created any grammar , i'am passing lm and dict created using lmtool as parameters to pocketsphinx . I have only 33 words to be recognised for my application .
I'll try -beam and -wbeam to make it faster. Also one more question is it possible to adapt the caosutic model on ubuntu running on VM on my windows m/c and then use it on raspberry pi ? or the adapatation needs to be done on raspberry itself?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Getting an error "Input overrun, read calls are too rare(non fatal)" . Mic settings are fine , and i have given the lm and dict on the command line. How do i get rid of this error?
Don't use very large grammar otherwise recognizer can not process incoming speech in time. Probably you also need to optimize recognizer to run faster on your platform. You can play with -beam and -wbeam parameters to make it faster.
I have'nt created any grammar , i'am passing lm and dict created using lmtool as parameters to pocketsphinx . I have only 33 words to be recognised for my application .
I'll try -beam and -wbeam to make it faster. Also one more question is it possible to adapt the caosutic model on ubuntu running on VM on my windows m/c and then use it on raspberry pi ? or the adapatation needs to be done on raspberry itself?