|
SPRAAK
|
All Data Structures Namespaces Files Functions Variables Typedefs Enumerations Enumerator Groups Pages
|
SPRAAK
|
Initialize an hmm file on the basis of unit or state level segmentations. More...
Functions | |
| SprHmmHmm * | spr_segpass_do (const SprCorSegOD *dbase, const SprHmmReadOD *acmod, const SprHmmSetupOD *eval) |
Initialize an hmm file on the basis of unit or state level segmentations.
spr_segpass [-ct FileName] <-c Corpus> [-range b:e](0:-1) [-ssp script](SPR_BSS_DEV_NULL) [-obs ObsDir] [-suffix FileSuffix](sam) [-arcd FileName] <-h FileName> [-g FileName] [-sel FileName] [-am_opt Options] [-top_n Number(s)](0) [-rmg rmg_params](no) [-LMout Value](-100) [-NOGS](flag: no gauss sel) [-ho FileName] [-go FileName] [-selo FileName] [-MinCount Value](1.0) [-LMsil Value](-10.0) [-LMspch Value](-10.0) [-REP](flag: relative minimal probabilities) [-LMtrans Value](-10.0) [-sf Factor](1.0) [-pool PoolingString](no) [-vm Value](0.0) (-u==-arcd)
| -ct<em>FileName</em><a | name="spr_segpass.ct" class="el"> Write out the accumulated statistics. If specified, and no dedicated output models (options -ho, -go, -selo) are given, only the statistics will be written out and no new acoustic model will be created/written. |
| -c<em>Corpus</em><a | name="spr_segpass.c" class="el"> File with corpus entries or segmentations. |
| -range<em>b:e</em><a | name="spr_segpass.range" class="el"> Optional begin and end entry the corpus/segmentation file. Counting starts at 0. |
| -ssp<em>script</em><a | name="spr_segpass.ssp" class="el"> The signal processing script used to preprocess the input data. |
| -obs<em>ObsDir</em><a | name="spr_segpass.obs" class="el"> Observation directory name. |
| -suffix<em>FileSuffix</em><a | name="spr_segpass.suffix" class="el"> File suffix of the observation files (without leading '.'). |
| -arcd<em>FileName</em><a | name="spr_segpass.arcd" class="el"> Unit file name (.arcd or .cd format). |
| -h<em>FileName</em><a | name="spr_segpass.h" class="el"> The input HMM file. |
| -g<em>FileName</em><a | name="spr_segpass.g" class="el"> The input MVG file (gaussians). |
| -sel<em>FileName</em><a | name="spr_segpass.sel" class="el"> The input select file name (tied gaussian). |
| -am_opt<em>Options</em><a | name="spr_segpass.am_opt" class="el"> Extra options for loading the acoustic model. A non-default acoustic model can be selected by having '=<am_type>;' as first option. See cwr_am_tbl.c for a list of acoustic models available. |
| -top_n<em>Number(s)</em><a | name="spr_segpass.top_n" class="el"> Only take the top-N gaussians into account when calculating output probabilities. If one value is given, it is used for all mixtures. Else a value per mixture must be given, separated by commas. Use '0' to set top_n to the number of gaussians in the mixture. |
| -rmg<em>rmg_params</em><a | name="spr_segpass.rmg" class="el"> The parameters for the quick selection of gaussians. If one value is given, it is used for all mixtures. Else a value per mixture must be given, separated by commas. Use 'no' if no quick selection is wanted. See rm_gauss.c for a description of the parameters. |
| -LMout<em>Value</em><a | name="spr_segpass.LMout" class="el"> Floor the state likelihoods of an observation using a fraction of the unconditional likelihood of the observation (weighted sum of the state likelihoods). Practically necessary if only few gaussians are evalutated (-top_n or -rmg options). The value given offset an automatically determined log10(fraction). Use -100 to turn the flooring off, and 0.0 to use the default.automatically. |
| -NOGSflag | no gauss sel Forgo the (sentence level) lexicon based Gaussian selection. The lexicon based Gaussian selection speeds up the decoding but may interfere with score normalization techniques that assume all Gaussians were evaluated. |
| -ho<em>FileName</em><a | name="spr_segpass.ho" class="el"> The output HMM file. Equals the input HMM if not specified (and no count file is requested). |
| -go<em>FileName</em><a | name="spr_segpass.go" class="el"> The output MVG file. Equals the input MVG if not specified (and no count file is requested). |
| -selo<em>FileName</em><a | name="spr_segpass.selo" class="el"> The output select file name (tied gaussian). |
| -MinCount<em>Value</em><a | name="spr_segpass.MinCount" class="el"> Minimal count required for the state statistics to be updated. |
| -LMsil<em>Value</em><a | name="spr_segpass.LMsil" class="el"> Minimal (logarithmic) probability for silence. |
| -LMspch<em>Value</em><a | name="spr_segpass.LMspch" class="el"> Minimal (logarithmic) probability for speech. |
| -REPflag | relative minimal probabilities For each label probability / gaussian weight, a different minimal value is calculated from the value given with the option -LMspch, proportional to its relative occurrence in the data. For explanation, see ct_to_oprob2(). |
| -LMtrans<em>Value</em><a | name="spr_segpass.LMtrans" class="el"> Minimal (logarithmic) probability for transition. |
| -sf<em>Factor</em><a | name="spr_segpass.sf" class="el"> Smoothing factor: smooth the output HMM with the input HMM (mode=TRAIN), or with uniform probabilities (mode=INIT). |
| -pool<em>PoolingString</em><a | name="spr_segpass.pool" class="el"> P: Pooled: ovlen dimensional sigma vector is the same for all gaussians. FP: Fully Pooled: only one sigma remains, it is the same for all gaussians AND dimensions. |
| -vm<em>Value</em><a | name="spr_segpass.vm" class="el"> Lower limit on the sqrt(variance) relative w.r.t. the weighted average of the variances. |
Train an hmm file on the basis of unit or state level segmentations. Initialises an hmm file on the basis of unit or state level segmentations.
The segmentations are read from the given segmenation file.
The frames in the observations are divided uniformly over the states in each unit.
For all continuous initializations, a correct mvg-file must be provided (e.g. produced by mvg_init -VC -i data -o mvgs -nc nr_mvgs). During the initialization, the gaussians are reestimated (based on the input gaussians).
If the timebase used in the corpus file is discrete, the frameshift is searched for in the corpus header. In the continuous case, the frameshift found in the data files will be used. The default the frame shift is 10 msec.
1.8.6