[hts-users:03434] Re: Issue with GV

[Keiichiro Oura <uratec@xxxxxxxxxxxxxxx>]

Hi,

You can try to change many settings for GV in the HTS demo scripts.

  MAXGVITER   maximum number of iterations of GV-based parameter
generation algorithm (default=50)
  GVEPSILON   convergence factor for GV iteration (default=0.0001)
  MINEUCNORM  minimum Euclid norm for GV iteration (default=0.01)
  STEPINIT    initial step size (default=1.0)
  STEPINC     step size acceleration factor (default=1.2)
  STEPDEC     step size deceleration factor (default=0.5)
  HMMWEIGHT   weight for HMM output prob. (default=1.0)
  GVWEIGHT    weight for GV output prob. (default=1.0)
  OPTKIND     optimization method (STEEPEST, NEWTON, or LBFGS) (default=NEWTON)
  NOSILGV     turn on GV without silent and pause phoneme (0:off or
1:on, default=1)
  CDGV        turn on context-dependent GV (0:off or 1:on, default=1)

Regards,
Keiichiro Oura


2012/10/19 Veera Raghavendra <raghavendra@xxxxxxxxxx>:
> Dear All,
>
> I found the difference in the quality with and without GV.
>
> If GV is turned-on, the voice quality is very clear but there are sudden
> spikes. These spikes destroys the synthesis quality.
>
> If GV is turned-off, there are no spikes but the content is not clear.
>
> Do I need to change any settings in HSMMAlign.
>
> Thanks,
> Raghavendra.