Can I use PLP features with sphinx3?

If there are the lowest-value and upper-value with cepstra in sphinx3 limited?

--- (Edited on 12/9/2008 1:08 am [GMT-0600] by chn) ---

> Can I use PLP features with sphinx3?

Yes

> If there are the lowest-value and upper-value with cepstra in sphinx3 limited?

No, there is only cmninit parameter you need to adjust if you are using live mode.

--- (Edited on 12/9/2008 1:25 am [GMT-0600] by nsh) ---

It's my feature file,which be converted to text file

740

0.000000

0.000000

0.000000

0.000000

0.000000

0.000000

0.000000

0.000000

0.000000

380.680756

400.391876

400.220459

399.998169

399.841370

399.707367

399.562073

421.367798

421.205719

421.070404

420.951263

420.961945

421.030396

0.000000

444.496033

444.279358

444.188110

444.156189

444.055084

443.979980

444.051910

444.072784

444.109955

444.064178

444.078674

444.100250

444.139008

444.186249

.........

dimention is 1,with batch mode no live mode.

I trained with it and with no any errors in my trainning.

But it's fatal errors when I decode it with sphinx3 .

And if I set cmn_arg with "none",as if decoder will no execu without cmn? 

--- (Edited on 12/9/2008 11:01 pm [GMT-0600] by chn) ---

--- (Edited on 12/9/2008 11:13 pm [GMT-0600] by chn) ---

It's quite strange decision to use dimension 1. The usual feature dimension is from 15 to 50.

--- (Edited on 12/10/2008 2:56 am [GMT-0600] by nsh) ---

I know , I compare it with another features!

--- (Edited on 12/10/2008 8:26 am [GMT-0600] by chn) ---

I want to use Sphinx3 with a different set of features, of size 80. I want to know whether Sphinx3 and SphinTrain support using these many features

--- (Edited on 11/20/2012 6:03 am [GMT-0600] by Visitor) ---

> I want to use Sphinx3 with a different set of features, of size 80. I want to know whether Sphinx3 and SphinTrain support using these many features

--- (Edited on 11/20/2012 17:17 [GMT+0300] by nsh) ---