General Discussion

Nested
FLAT START MONOPHONES: Step#6 Issue ..!!
User: linx
Date: 12/3/2014 3:34 am
Views: 2276
Rating: 1

All,

I am newbie and following Voxforge STEP#6 link - http://www.voxforge.org/home/dev/acousticmodels/linux/create/htkjulius/tutorial/monophones/step-6 for creating FLAT START MONOPHONES.

I tried below command but getting errors as -

---

$ HERest -A -D -T 1 -C config -I phones0.mlf -t 250.0 150.0 1000.0 -S train.scp -H hmm0/macros -H hmm0/hmmdefs -M hmm1 monophones0
HERest -A -D -T 1 -C config -I phones0.mlf -t 250.0 150.0 1000.0 -S train.scp -H hmm0/macros -H hmm0/hmmdefs -M hmm1 monophones0

HTK Configuration Parameters[10]
  Module/Tool     Parameter                  Value
#                 NUMCEPS                       12
#                 CEPLIFTER                     22
#                 NUMCHANS                      26
#                 PREEMCOEF               0.970000
#                 USEHAMMING                  TRUE
#                 WINDOWSIZE         250000.000000
#                 SAVEWITHCRC                 TRUE
#                 SAVECOMPRESSED              TRUE
#                 TARGETRATE         100000.000000
#                 TARGETKIND          MFCC_0_D_N_Z

HERest  ML Updating: Transitions Means Variances

 System is PLAIN
38 Logical/38 Physical Models Loaded, VecSize=25
2 MMF input files
Pruning-On[250.0 150.0 1000.0]
 Processing Data: cima1.mfc; Label X1.lab
  ERROR [+6510]  LOpen: Unable to open label file ../train/mfcc/X1.lab
 FATAL ERROR - Terminating program HERest
---

But, if I HACK as -

“cp ../../manual/phones0.mlf .” to /train/mfcc folder, rename as .lab file. Delete first two lines, last '.' line and remove 'sil' from .lab file. Huuuuh..  it works as -

---

$ HERest -A -D -T 1 -C config -I phones0.mlf -t 250.0 150.0 1000.0 -S train.scp -H hmm0/macros -H hmm0/hmmdefs -M hmm1 monophones0
HERest -A -D -T 1 -C config -I phones0.mlf -t 250.0 150.0 1000.0 -S train.scp -H hmm0/macros -H hmm0/hmmdefs -M hmm1 monophones0

HTK Configuration Parameters[10]
  Module/Tool     Parameter                  Value
#                 NUMCEPS                       12
#                 CEPLIFTER                     22
#                 NUMCHANS                      26
#                 PREEMCOEF               0.970000
#                 USEHAMMING                  TRUE
#                 WINDOWSIZE         250000.000000
#                 SAVEWITHCRC                 TRUE
#                 SAVECOMPRESSED              TRUE
#                 TARGETRATE         100000.000000
#                 TARGETKIND          MFCC_0_D_N_Z

HERest  ML Updating: Transitions Means Variances

 System is PLAIN
38 Logical/38 Physical Models Loaded, VecSize=25
2 MMF input files
Pruning-On[250.0 150.0 1000.0]
 Processing Data: X1.mfc; Label X1.lab
 Utterance prob per frame = -6.575663e+01
 WARNING [-2331]  UpdateModels: a[1] copied: only 1 egs
 in HERest
 WARNING [-2331]  UpdateModels: g[5] copied: only 1 egs
 in HERest
 WARNING [-2331]  UpdateModels: v[14] copied: only 2 egs
 in HERest
 WARNING [-2331]  UpdateModels: aw[22] copied: only 1 egs
 in HERest
 WARNING [-2331]  UpdateModels: en[28] copied: only 1 egs
 in HERest
 WARNING [-2331]  UpdateModels: sh[36] copied: only 1 egs
 in HERest
Saving hmm's to dir hmm1
Reestimation complete - average log prob per frame = -6.575663e+01
     - total frames seen          = 5.998000e+03

HTK Configuration Parameters[10]
  Module/Tool     Parameter                  Value
                  NUMCEPS                       12
                  CEPLIFTER                     22
                  NUMCHANS                      26
                  PREEMCOEF               0.970000
                  USEHAMMING                  TRUE
                  WINDOWSIZE         250000.000000
                  SAVEWITHCRC                 TRUE
                  SAVECOMPRESSED              TRUE
                  TARGETRATE         100000.000000
                  TARGETKIND          MFCC_0_D_N_Z

---


I know the way I HACK is NOT proper way.

Please suggest. Thanks ..!!

--- (Edited on 12/3/2014 3:34 am [GMT-0600] by ) ---

PreviousNext