Viterbi alignment + path probabilities. More...

Detailed Description

Viterbi alignment + path probabilities.

spr_vitprobs [-WM](flag: word markers) [-NN](flag: non-normalized likelihoods)
    [-GS](flag: select gaussians) [-seg segfname](stdout) <-beam beam search info> <-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)
    <-ci unit description files> [-cd unit description files] <-d dictionary>
    [-unwind unwind format] (-u==-ci)

Parameters

-WMflag	word markers Insert word markers in the segmentation file.
-NNflag	non-normalized likelihoods Output non-normalized likelihoods.
-GSflag	select gaussians Select the set of active Gaussians based on the content of each sentence (usually not a good idea if you want posterioirs).
-seg<em>segfname</em><a	name="spr_vitprobs.seg" class="el"> File to write the segmentations to.
-beam<em>beam	search info A beam search algorithm will be used. The extra parameters that can be specified are explained in the manual page of load_search_space().
-c<em>Corpus</em><a	name="spr_vitprobs.c" class="el"> File with corpus entries or segmentations.
-range<em>b:e</em><a	name="spr_vitprobs.range" class="el"> Optional begin and end entry the corpus/segmentation file. Counting starts at 0.
-ssp<em>script</em><a	name="spr_vitprobs.ssp" class="el"> The signal processing script used to preprocess the input data.
-obs<em>ObsDir</em><a	name="spr_vitprobs.obs" class="el"> Observation directory name.
-suffix<em>FileSuffix</em><a	name="spr_vitprobs.suffix" class="el"> File suffix of the observation files (without leading '.').
-arcd<em>FileName</em><a	name="spr_vitprobs.arcd" class="el"> Unit file name (.arcd or .cd format).
-h<em>FileName</em><a	name="spr_vitprobs.h" class="el"> The input HMM file.
-g<em>FileName</em><a	name="spr_vitprobs.g" class="el"> The input MVG file (gaussians).
-sel<em>FileName</em><a	name="spr_vitprobs.sel" class="el"> The input select file name (tied gaussian).
-am_opt<em>Options</em><a	name="spr_vitprobs.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_vitprobs.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_vitprobs.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_vitprobs.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.
-ci<em>unit	description files The two units description files seperated by white-space. The first file just lists the units (phones). The second file describes the context dependencies.
-cd<em>unit	description files The two units description files seperated by white-space. The first file just lists the units (phones). The second file describes the context dependencies.
-d<em>dictionary</em><a	name="spr_vitprobs.d" class="el"> Dictionary file name.
-unwind<em>unwind	format Define the parameters to modify the word transcriptions. See spr_cwr_lex_desc_read() for more details.

Viterbi alignment + path probabilities. Viterbi alignment of all entries in a corpus file against the correct model (spelling entry in the corpus file). The path (state level) and the normalized scores per frame are written in text format to a segmentation file. The format of each line is:

<fname> <state> <f0> <nfr> <l_0> ... <l_(nfr-1)> <t0> <t1>

With fname the file name or '-' when the segment belongs to the same entry as the previous segment. Next comes the HMM-state described as a the sequence u#nr, with u the unit name (phoneme), and nr the state number (0...Nstate-1). Next come the first frame index f0 and the number of frames nfr assigned to this state. These first four fields just make up a standard segmentation record. The next elements are the normalize frame likelihoods l_i

l_i = f(x|<u>#<nr>)/f(x)

for each frame. f(x) is esimated as:

sum(f(x|s) P(s)), with P(s) the unconditional state probabilities.

And finaly the self t0 and exit t1 transition probabilities are listed. Note that all probabilities are log10() values.

Author: Kris Demuynck

Date: 23/02/01