|
'KALDI'
|
|
'KALDI'
|
This page contains a list of all the Kaldi tools, with their brief functions and usage messages.
Tools | Description |
| align-equal.cc | Write equally spaced alignments of utterances (to get training started) Usage: align-equal <tree-in> <model-in> <lexicon-fst-in> <features-rspecifier> <transcriptions-rspecifier> <alignments-wspecifier> e.g.: align-equal 1.tree 1.mdl lex.fst scp:train.scp ark:train.tra ark:equal.ali |
| align-equal-compiled.cc | Write an equally spaced alignment (for getting training started)Usage: align-equal-compiled <graphs-rspecifier> <features-rspecifier> <alignments-wspecifier> e.g.: align-equal-compiled 1.mdl 1.fsts scp:train.scp ark:equal.ali |
| acc-tree-stats.cc | Accumulate statistics for phonetic-context tree building. Usage: acc-tree-stats [options] model-in features-rspecifier alignments-rspecifier [tree-accs-out] e.g.: acc-tree-stats 1.mdl scp:train.scp ark:1.ali 1.tacc |
| show-alignments.cc | Display alignments in human-readable form Usage: show-alignments [options] <phone-syms> <model> <alignments-rspecifier> e.g.: show-alignments phones.txt 1.mdl ark:1.ali |
| compile-questions.cc | Compile questions Usage: compile-questions [options] <topo> <questions-text-file> <questions-out> e.g.: compile-questions questions.txt questions.qst |
| cluster-phones.cc | Cluster phones (or sets of phones) into sets for various purposes Usage: cluster-phones [options] <tree-stats-in> <phone-sets-in> <clustered-phones-out> e.g.: cluster-phones 1.tacc phonesets.txt questions.txt |
| compute-wer.cc | Compute WER by comparing different transcriptions Takes two transcription files, in kaldi integer format Usage: compute-wer [options] <ref-rspecifier> <hyp-rspecifier> |
| make-h-transducer.cc | Make H transducer from transition-ids to context-dependent phones, without self-loops [use add-self-loops to add them] Usage: make-h-transducer ilabel-info-file tree-file transition-gmm/acoustic-model [H-fst-out] e.g.: make-h-transducer ilabel_info 1.tree 1.mdl > H.fst |
| add-self-loops.cc | Add self-loops and transition probabilities to transducer, expanding to transition-ids Usage: add-self-loops [options] transition-gmm/acoustic-model [fst-in] [fst-out] e.g.: add-self-loops --self-loop-scale=0.1 1.mdl < HCLG_noloops.fst > HCLG_full.fst |
| convert-ali.cc | Convert alignments from one decision-tree/model to another Usage: convert-ali [options] old-model new-model new-tree old-alignments-rspecifier new-alignments-wspecifier e.g.: convert-ali old.mdl new.mdl new.tree ark:old.ali ark:new.ali |
| compile-train-graphs.cc | Creates training graphs (without transition-probabilities, by default) Usage: compile-train-graphs [options] tree-in model-in lexicon-fst-in transcriptions-rspecifier graphs-wspecifier e.g.: compile-train-graphs tree 1.mdl lex.fst ark:train.tra ark:graphs.fsts |
| compile-train-graphs-fsts.cc | Creates training graphs (without transition-probabilities, by default) This version takes FSTs as inputs (e.g., representing a separate weighted grammar for each utterance) Note: the lexicon should contain disambiguation symbols and you should supply the --read-disambig-syms option which is the filename of a list of disambiguation symbols. Warning: you probably want to set the --transition-scale and --self-loop-scale options; the defaults (zero) are probably not appropriate. Usage: compile-train-graphs-fsts [options] tree-in model-in lexicon-fst-in graphs-rspecifier graphs-wspecifier e.g.: compile-train-graphs-fsts --read-disambig-syms=disambig.list\ tree 1.mdl lex.fst ark:train.fsts ark:graphs.fsts |
| make-pdf-to-tid-transducer.cc | Make transducer from pdfs to transition-ids Usage: make-pdf-to-tid-transducer model-filename [fst-out] e.g.: make-pdf-to-tid-transducer 1.mdl > pdf2tid.fst |
| make-ilabel-transducer.cc | Make transducer that de-duplicates context-dependent ilabels that map to the same state Usage: make-ilabel-transducer ilabel-info-right tree-file transition-gmm/model ilabel-info-left [mapping-fst-out] e.g.: make-ilabel-transducer old_ilabel_info 1.tree 1.mdl new_ilabel_info > convert.fst |
| show-transitions.cc | Print debugging info from transition model Usage: show-transitions phones-symbol-table transition/model-file [occs-file] e.g.: show-transitions phones.txt 1.mdl 1.occs |
| ali-to-phones.cc | Convert model-level alignments to phone-sequences (in integer, not text, form) Usage: ali-to-phones [options] <model> <alignments-rspecifier> <phone-transcript-wspecifier> e.g.: ali-to-phones 1.mdl ark:1.ali ark:phones.tra |
| ali-to-post.cc | Convert alignments to posteriors Usage: ali-to-post [options] alignments-rspecifier posteriors-wspecifier e.g.: ali-to-post ark:1.ali ark:1.post |
| weight-silence-post.cc | Apply weight to silences in posteriors Usage: weight-silence-post [options] <silence-weight> <silence-phones> <model> <posteriors-rspecifier> <posteriors-wspecifier> e.g.: weight-silence-post 0.0 1:2:3 1.mdl ark:1.post ark:nosil.post |
| acc-lda.cc | Accumulate LDA statistics based on pdf-ids. Usage: acc-lda [options] <transition-gmm/model> <features-rspecifier> <posteriors-rspecifier> <lda-acc-out> Typical usage: ali-to-post ark:1.ali ark:- | lda-acc 1.mdl "ark:splice-feats scp:train.scp|" ark:- ldaacc.1 |
| est-lda.cc | Estimate LDA transform using stats obtained with acc-lda. Usage: est-lda [options] <lda-matrix-out> <lda-acc-1> <lda-acc-2> ... |
| ali-to-pdf.cc | Converts alignments (containing transition-ids) to pdf-ids, zero-based. Usage: ali-to-pdf [options] <model> <alignments-rspecifier> <pdfs-wspecifier> e.g.: ali-to-pdf 1.mdl ark:1.ali ark, t:- |
| est-mllt.cc | Do MLLT update Usage: est-mllt [options] <mllt-mat-out> <stats-in1> <stats-in2> ... e.g.: est-mllt 2.mat 1a.macc 1b.macc ... Note: use compose-transforms <mllt-mat-out> <prev-mllt-mat> to combine with previous MLLT or LDA transform, if any, and gmm-transform-means to apply <mllt-mat-out> to GMM means. |
| build-tree.cc | Train decision tree Usage: build-tree [options] <tree-stats-in> <roots-file> <questions-file> <topo-file> <tree-out> e.g.: build-tree treeacc roots.txt 1.qst topo tree |
| build-tree-two-level.cc | Trains two-level decision tree. Outputs the larger tree, and a mapping from the leaf-ids of the larger tree to those of the smaller tree. Useful, for instance, in tied-mixture systems with multiple codebooks. Usage: build-tree-two-level [options] <tree-stats-in> <roots-file> <questions-file> <topo-file> <tree-out> <mapping-out> e.g.: build-tree-two-level treeacc roots.txt 1.qst topo tree tree.map |
| decode-faster.cc | Decode, reading log-likelihoods (of transition-ids or whatever symbol is on the graph) as matrices Usage: decode-faster [options] fst-in loglikes-rspecifier words-wspecifier [alignments-wspecifier] |
| decode-faster-mapped.cc | Decode, reading log-likelihoods as matrices (model is needed only for the integer mappings in its transition-model) Usage: decode-faster-mapped [options] model-in fst-in loglikes-rspecifier words-wspecifier [alignments-wspecifier] |
| scale-vecs.cc | Scale a set of vectors in a Table (useful for speaker vectors) Usage: scale-vecs [options] in-rspecifier out-wspecifier |
| copy-transition-model.cc | Copies a transition model (this can be used to separate transition models from the acoustic models they are written with. Usage: show-transitions [options] <transition-model or model file> <transition-model-out> e.g.: show-transitions phones.txt 1.mdl 1.acc |
| rand-prune-post.cc | Randomized pruning of posteriors less than threshold Note: for posteriors derived from alignments, threshold must be greater than one, or this will have no effect (speedup factor will be roughly the same as the threshold) Usage: rand-prune-post [options] <rand-prune-value> <posteriors-rspecifier> <posteriors-wspecifier> e.g.: rand-prune-post 5.0 ark:- ark:- |
| phones-to-prons.cc | Convert pairs of (phone-level, word-level) transcriptions to output that indicates the phones assigned to each word. Format is standard format for archives of vector<vector<int32> > i.e. : utt-id 600 4 7 19 ; 512 4 18 ; 0 1 where 600, 512 and 0 are the word-ids (0 for non-word phones, e.g. optional-silence introduced by the lexicon), and the phone-ids follow the word-ids. Note: L_align.fst must have word-start and word-end symbols in it Usage: phones-to-prons [options] <L_align.fst> <word-start-sym> <word-end-sym> <phones-rspecifier> <words-rspecifier> <prons-wspecifier> e.g.: ali-to-phones 1.mdl ark:1.ali ark:- | \ phones-to-prons L_align.fst 46 47 ark:- 1.tra ark:1.prons |
| prons-to-wordali.cc | Given per-utterance pronunciation information as output by words-to-prons, and per-utterance phone alignment information as output by ali-to-phones --write-lengths, output word alignment information that can be turned into the ctm format. Outputs is pairs of (word, #frames), or if --per-frame is given, just the word for each frame. Note: zero word-id usually means optional silence. Format is standard format for archives of vector<pair<int32, int32> > i.e. : utt-id 600 22 ; 1028 32 ; 0 41 where 600, 1028 and 0 are the word-ids, and 22, 32 and 41 are the lengths. Usage: prons-to-wordali [options] <prons-rspecifier> <phone-lengths-rspecifier> <wordali-wspecifier> e.g.: ali-to-phones 1.mdl ark:1.ali ark:- | \ phones-to-prons L_align.fst 46 47 ark:- 1.tra ark:- | \ prons-to-wordali ark:- \ "ark:ali-to-phones --write-lengths 1.mdl ark:1.ali ark:-|" ark:1.wali |
| copy-gselect.cc | Copy Gaussian indices for pruning, possibly making the lists shorter (e.g. the --n=10 limits to the 10 best indices Usage: gmm-gselect [options] <gselect-rspecifier> <gselect-wspecifier> |
| copy-tree.cc | Copy decision tree (possibly changing binary/text format) Usage: copy-tree [--binary=false] <tree-in> <tree-out> |
| scale-post.cc | Scale posteriors with either a global scale, or a different scale for each utterance. Usage: scale-post post-rspecifier (scale-rspecifier|scale) post-wspecifier |
| compute-mce-scale.cc | compute the scale of MCE, which is used to scale posteriors Usage: compute-mce-scale [option] num-score-rspecifier den-score-rspecifier out-scale-wspecifier |
| get-silence-probs.cc | This program takes two archives of Vector<BaseFloat>, representing per-frame log-likelihoods for silence and non-silence models respectively. It outputs per-frame silence probabilities in the same format. To get non-silence probs instead, use --write-nonsil-probs Usage: get-silence-probs [options] <silence-loglikes-rspecifier> <nonsilence-loglikes-rspecifier> <silence-probs-wspecifier> e.g.: get-silence-probs --silence-prior=0.9 --quantize=0.25 ark:sil.likes ark:nonsil.likes ark:sil.probs |
| post-to-weights.cc | Turn posteriors into per-frame weights (typically most useful after weight-silence-post, to get silence weights) Usage: post-to-weights post-rspecifier weights-wspecifier |
| reverse-weights.cc | Modify per-frame weights by outputting 1.0-weight (if --reverse=true); if --reverse=false, do nothing to them. Usage: reverse-weights weights-rspecifier weights-wspecifier |
| dot-weights.cc | Takes two archives of vectors (typically representing per-frame weights) and for each utterance, outputs the dot product. Useful for evaluating the accuracy of silence classifiers. Usage: dot-weights weights-rspecifier1 weights-rspecifier2 float-wspecifier |
| sum-tree-stats.cc | Sum statistics for phonetic-context tree building. Usage: sum-tree-stats [options] tree-accs-out tree-accs-in1 tree-accs-in2 ... e.g.: sum-tree-stats treeacc 1.treeacc 2.treeacc 3.treeacc |
| weight-post.cc | Takes archives (typically per-utterance) of posteriors and per-frame weights, and weights the posteriors by the per-frame weights Usage: weight-post <post-rspecifier> <weights-rspecifier> <post-wspecifier> |
| post-to-tacc.cc | From posteriors, compute transition-accumulators The output is a vector of counts/soft-counts, indexed by transition-id) Note: the model is only read in order to get the size of the vector Usage: post-to-tacc [options] <model> <post-rspecifier> <accs> e.g.: post-to-tacc --binary=false 1.mdl "ark:ali-to-post 1.ali|" 1.tacc Useful options include the --weights option |
| copy-matrix.cc | Copy matrices, or archives of matrices (e.g. features or transforms) Usage: copy-matrix [options] (matrix-in-rspecifier|matrix-in-rxfilename) (matrix-out-wspecifier|matrix-out-wxfilename) e.g.: copy-matrix --binary=false 1.mat - copy-matrix ark:2.trans ark,t:- |
| copy-vector.cc | Copy vectors, or archives of vectors (e.g. transition-accs; speaker vectors) Usage: copy-vector [options] (vector-in-rspecifier|vector-in-rxfilename) (vector-out-wspecifier|vector-out-wxfilename) e.g.: copy-vector --binary=false 1.mat - copy-vector ark:2.trans ark,t:- |
| copy-int-vector.cc | Copy vectors of integers, or archives of vectors of integers (e.g. alignments) Usage: copy-int-vector [options] (vector-in-rspecifier|vector-in-rxfilename) (vector-out-wspecifier|vector-out-wxfilename) e.g.: copy-int-vector --binary=false foo - copy-int-vector ark:1.ali ark,t:- |
| sum-post.cc | Sum two sets of posteriors for each utterance, e.g. useful in fMMI. To take the difference of posteriors, use e.g. --scale2=-1.0 Usage: sum-post post-rspecifier1 post-rspecifier2 post-wspecifier |
| sum-matrices.cc | Sum matrices, e.g. stats for fMPE training Usage: sum-matrices [options] <mat-out> <mat-in1> <mat-in2> ... e.g.: sum-matrices mat 1.mat 2.mat 3.mat |
| draw-tree.cc | Outputs a decision tree description in GraphViz format Usage: draw-tree [options] <phone-symbols> <tree> e.g.: draw-tree phones.txt tree | dot -Gsize=8,10.5 -Tps | ps2pdf - tree.pdf |
| copy-int-vector-vector.cc | Copy vectors of vectors of integers, or archives thereof Usage: copy-int-vector-vector [options] vector-in-(rspecifier|rxfilename) vector-out-(wspecifierwxfilename) |
| align-mapped.cc | Generate alignments, reading log-likelihoods as matrices. (model is needed only for the integer mappings in its transition-model) Usage: align-mapped [options] tree-in trans-model-in lexicon-fst-in feature-rspecifier transcriptions-rspecifier alignments-wspecifier e.g.: gmm-align tree trans.mdl lex.fst scp:train.scp ark:train.tra ark:nnet.ali |
| align-compiled-mapped.cc | Generate alignments, reading log-likelihoods as matrices. (model is needed only for the integer mappings in its transition-model) Usage: align-compiled-mapped [options] trans-model-in graphs-rspecifier feature-rspecifier alignments-wspecifier e.g.: nnet-align-compiled trans.mdl ark:graphs.fsts scp:train.scp ark:nnet.ali or: compile-train-graphs tree trans.mdl lex.fst ark:train.tra b, ark:- | \ nnet-align-compiled trans.mdl ark:- scp:loglikes.scp t, ark:nnet.ali |
| latgen-faster-mapped.cc | Generate lattices, reading log-likelihoods as matrices (model is needed only for the integer mappings in its transition-model) Usage: latgen-faster-mapped [options] trans-model-in (fst-in|fsts-rspecifier) loglikes-rspecifier lattice-wspecifier [ words-wspecifier [alignments-wspecifier] ] |
| compute-mfcc-feats.cc | Create MFCC feature files. Usage: compute-mfcc-feats [options...] <wav-rspecifier> <feats-wspecifier> |
| compute-plp-feats.cc | Create PLP feature files. Usage: compute-plp-feats [options...] <wav-rspecifier> <feats-wspecifier> |
| compute-fbank-feats.cc | Create Mel-filter bank (FBANK) feature files. Usage: compute-fbank-feats [options...] <wav-rspecifier> <feats-wspecifier> |
| compute-cmvn-stats.cc | Compute cepstral mean and variance normalization statistics If wspecifier provided: per-utterance by default, or per-speaker if spk2utt option provided; if wxfilename: global Usage: compute-cmvn-stats [options] feats-rspecifier (stats-wspecifier|stats-wxfilename) |
| add-deltas.cc | Add deltas (typically to raw mfcc or plp features Usage: add-deltas [options] in-rspecifier out-wspecifier |
| remove-mean.cc | Remove mean from each feature file [ for per-speaker normalization, use add-cmvn-stats and apply-cmvn ] Usage: remove-mean [options] in-rspecifier out-wspecifier |
| apply-cmvn.cc | Apply cepstral mean and (optionally) variance normalization Per-utterance by default, or per-speaker if utt2spk option provided Usage: apply-cmvn [options] (cmvn-stats-rspecifier|cmvn-stats-rxfilename) feats-rspecifier feats-wspecifier |
| transform-feats.cc | Apply transform (e.g. LDA; HLDA; fMLLR/CMLLR; MLLT/STC) Linear transform if transform-num-cols == feature-dim, affine if transform-num-cols == feature-dim+1 (->append 1.0 to features) Per-utterance by default, or per-speaker if utt2spk option provided Global if transform-rxfilename provided. Usage: transform-feats [options] (transform-rspecifier|transform-rxfilename) feats-rspecifier feats-wspecifier |
| copy-feats.cc | Copy features [and possibly change format] Usage: copy-feats [options] in-rspecifier out-wspecifier |
| compose-transforms.cc | Compose (affine or linear) feature transforms Usage: compose-transforms [options] (transform-A-rspecifier|transform-A-rxfilename) (transform-B-rspecifier|transform-B-rxfilename) (transform-out-wspecifier|transform-out-wxfilename) Note: it does matrix multiplication (A B) so B is the transform that gets applied to the features first. If b-is-affine = true, then assume last column of b corresponds to offset e.g.: compose-transforms 1.mat 2.mat 3.mat compose-transforms 1.mat ark:2.trans ark:3.trans compose-transforms ark:1.trans ark:2.trans ark:3.trans |
| splice-feats.cc | Splice features with left and right context (e.g. prior to LDA) Usage: splice-feats [options] in-rspecifier out-wspecifier |
| extract-segments.cc | Create MFCC feature files. Usage: extract-segments [options...] <wav-rspecifier> <segments-file> <wav-wspecifier> (segments-file has lines like: spkabc_seg1 spkabc_recording1 1.10 2.36 1 or: spkabc_seg1 spkabc_recording1 1.10 2.36 [if channel not provided as last element, expects mono.] |
| subset-feats.cc | Copy a subset of features Usage: subset-feats [options] in-rspecifier out-wspecifier |
| feat-to-len.cc | Reads an archive of features and writes a corresponding archive that maps utterance-id to utterance length in frames. Usage: feat-to-len [options] in-rspecifier out-wspecifier e.g.: feat-to-len scp:feats.scp ark,t:feats.lengths |
| feat-to-dim.cc | Reads an archive of features. If second argument is wxfilename, writes the feature dimension of the first feature file; if second argument is wspecifier, writes an archive of the feature dimension, indexed by utterance id. Usage: feat-to-dim [options] feat-rspecifier (dim-wspecifier|dim-wxfilename) e.g.: feat-to-dim scp:feats.scp - |
| fmpe-apply-transform.cc | Apply fMPE transform to features Usage: fmpe-apply-transform [options...] <fmpe-object> <feat-rspecifier> <gselect-rspecifier> <feat-wspecifier> |
| fmpe-acc-stats.cc | Compute statistics for fMPE training Usage: fmpe-acc-stats [options...] <fmpe-object> <feat-rspecifier> <feat-diff-rspecifier> <gselect-rspecifier> <stats-out> Note: gmm-fmpe-acc-stats avoids computing the features an extra time |
| fmpe-init.cc | Initialize fMPE transform (to zero) Usage: fmpe-init [options...] <diag-gmm-in> <fmpe-out> E.g. fmpe-init 1.ubm 1.fmpe |
| fmpe-est.cc | Initialize fMPE transform (to zeo) Usage: fmpe-est [options...] <fmpe-in> <stats-in> <fmpe-out> E.g. fmpe-est 1.fmpe 1.accs 2.fmpe |
| fmpe-copy.cc | Copy fMPE transform Usage: fmpe-init [options...] <fmpe-in> <fmpe-out> E.g. fmpe-copy --binary=false 1.fmpe text.fmpe |
| fgmm-global-acc-stats.cc | Accumulate stats for training a full-covariance GMM. Usage: fgmm-global-acc-stats [options] <model-in> <feature-rspecifier> <stats-out> e.g.: fgmm-global-acc-stats 1.mdl scp:train.scp 1.acc |
| fgmm-global-sum-accs.cc | Sum multiple accumulated stats files for full-covariance GMM training. Usage: fgmm-global-sum-accs [options] stats-out stats-in1 stats-in2 ... |
| fgmm-global-est.cc | Estimate a full-covariance GMM from the accumulated stats. Usage: fgmm-global-est [options] <model-in> <stats-in> <model-out> |
| fgmm-global-merge.cc | Combine a number of GMMs into a larger GMM, with #Gauss = sum(individual #Gauss)). Output full GMM, and a text file with sizes of each individual GMM. Usage: fgmm-global-merge [options] fgmm-out sizes-file-out fgmm-in1 fgmm-in2 ... |
| fgmm-global-to-gmm.cc | Convert single full-covariance GMM to single diagonal-covariance GMM. Usage: fgmm-global-to-gmm [options] 1.fgmm 1.gmm |
| fgmm-gselect.cc | Precompute Gaussian indices for pruning (e.g. in training UBMs, SGMMs, tied-mixture systems) For each frame, gives a list of the n best Gaussian indices, sorted from best to worst. Usage: fgmm-gselect [options] <model-in> <feature-rspecifier> <gselect-wspecifier> The --gselect option (which takes an rspecifier) limits selection to a subset of indices: e.g.: fgmm-gselect "--gselect=ark:gunzip -c bigger.gselect.gz|" --n=20 1.gmm "ark:feature-command |" "ark,t:|gzip -c >1.gselect.gz" |
| fgmm-global-get-frame-likes.cc | Print out per-frame log-likelihoods for each utterance, as an archive of vectors of floats. Usage: fgmm-global-get-frame-likes [options] <model-in> <feature-rspecifier> <likes-out-wspecifier> e.g.: fgmm-global-get-frame-likes 1.mdl scp:train.scp ark:1.likes |
| fgmm-global-acc-stats-twofeats.cc | Accumulate stats for training a full-covariance GMM, two-feature version Usage: fgmm-global-acc-stats-twofeats [options] <model-in> <feature1-rspecifier> <feature2-rspecifier> <stats-out> e.g.: fgmm-global-acc-stats-twofeats 1.mdl scp:train.scp scp:train2.scp 1.acc |
| fgmm-global-copy.cc | Copy a full-covariance GMM Usage: fgmm-global-copy [options] <model-in> <model-out> e.g.: fgmm-global-copy --binary=false 1.model - | less |
| fstpredeterminize.cc | Predeterminizes input FST by adding input symbols as necessary for fstdeterminizestar to succeed. Usage: fstpredeterminize disambig_out.list [in.fst [out.fst] ] |
| fstdeterminizestar.cc | Removes epsilons and determinizes in one step Usage: fstdeterminizestar [in.fst [out.fst] ] |
| fstrmsymbols.cc | Replaces a subset of symbols with epsilon, wherever they appear on the input side of an FST. Usage: fstrmsymbols in-disambig-list [in.fst [out.fst] ] E.g: fstrmsymbols in.list < in.fst > out.fst |
| fstisstochastic.cc | Checks whether an FST is stochastic and exits with success if so. Prints out maximum error (in log units). Usage: fstisstochastic [ in.fst ] |
| fstminimizeencoded.cc | Minimizes FST after encoding [similar to fstminimize, but no weight-pushing] Usage: fstminimizeencoded [in.fst [out.fst] ] |
| fstoptimize.cc | Optimizes FST using predeterminization, determinization, optional weight and symbol-pushing, and encoded minimization Usage: fstoptimize [in.fst [out.fst] ] |
| fstmakecontextfst.cc | Constructs a context FST with a specified context-width and context-position. Outputs the context FST, and a file in Kaldi format that describes what the input labels mean. Usage: fstmakecontextfst phones_symtab subseq_sym ilabels_output_file [out.fst] E.g.: fstmakecontextfst phones.txt 42 ilabels.sym > C.fst |
| fstmakecontextsyms.cc | Create input symbols for CLG Usage: fstmakecontextsyms phones-symtab ilabels_input_file [output-symtab.txt] E.g.: fstmakecontextsyms phones.txt ilabels.sym > context_symbols.txt |
| fstaddsubsequentialloop.cc | Minimizes FST after encoding [this algorithm applicable to all FSTs in tropical semiring] Usage: fstaddsubsequentialloop subseq_sym [in.fst [out.fst] ] E.g.: fstaddsubsequentialloop 52 < LG.fst > LG_sub.fst |
| fstaddselfloops.cc | Adds self-loops to states of an FST to propagate disambiguation symbols through it They are added on each final state and each state with non-epsilon output symbols on at least one arc out of the state. Useful in conjunction with predeterminize Usage: fstaddselfloops in-disambig-list out-disambig-list [in.fst [out.fst] ] E.g: fstaddselfloops in.list out.list < in.fst > withloops.fst |
| fstrmepslocal.cc | Removes some (but not all) epsilons in an algorithm that will always reduce the number of arcs+states. Option to preserves equivalence in tropical or log semiring, and if in tropical, stochasticit in either log or tropical. Usage: fstrmepslocal [in.fst [out.fst] ] |
| fstcomposecontext.cc | Composes on the left with a dynamically created context FST Usage: fstcomposecontext ilabels-output-file [in.fst [out.fst] ] E.g: fstcomposecontext ilabels.sym < LG.fst > CLG.fst |
| fsttablecompose.cc | Composition algorithm [between two FSTs of standard type, in tropical semiring] that is more efficient for certain cases-- in particular, where one of the FSTs (the left one, if --match-side=left) has large out-degree Usage: fsttablecompose (fst1-rxfilename|fst1-rspecifier) (fst2-rxfilename|fst2-rspecifier) [(out-rxfilename|out-rspecifier)] |
| fstrand.cc | Generate random FST Usage: fstrand [out.fst] |
| fstfactor.cc | Factor fst into two parts (by removing linear chains) Usage: fstfactor in.fst out1.fst out2.fst |
| fstdeterminizelog.cc | Determinizes in the log semiring Usage: fstdeterminizelog [in.fst [out.fst] ] |
| fstreorder.cc | Reorder FST states for greater search efficiency [sort arcs by weight then dfs order states] Usage: fstreorder [in.fst [out.fst] ] |
| fstremoveuselessarcs.cc | Remove useless arcs from an FST (those which will never be on a best path for any input sequence; also removes ties). Usage: fstremoveuselessarcs [in.fst [out.fst] ] |
| fstphicompose.cc | Composition, where the right FST has "failure" (phi) transitions that are only taken where there was no match of a "real" label You supply the label corresponding to phi. Usage: fstphicompose phi-label (fst1-rxfilename|fst1-rspecifier) (fst2-rxfilename|fst2-rspecifier) [(out-rxfilename|out-rspecifier)] E.g.: fstphicompose 54 a.fst b.fst c.fst or: fstphicompose 11 ark:a.fsts G.fst ark:b.fsts |
| fstpropfinal.cc | Propagates final-states through phi transitions Usage: fstpropfinal phi-label [in.fst [out.fst] ] |
| fstcopy.cc | Copy tables/archives of FSTs, index by utterance-id Usage: fstcopy <fst-rspecifier> <fst-wspecifier> |
| gmm-init-mono.cc | Initialize monophone GMM. Usage: gmm-init-mono <topology-in> <dim> <model-out> <tree-out> e.g.: gmm-init-mono topo 39 mono.mdl mono.tree |
| gmm-est.cc | Accumulate stats for GMM training. Usage: gmm-est [options] <model-in> <stats-in> <model-out> e.g.: gmm-est 1.mdl 1.acc 2.mdl |
| gmm-acc-stats-ali.cc | Accumulate stats for GMM training. Usage: gmm-acc-stats-ali [options] <model-in> <feature-rspecifier> <alignments-rspecifier> <stats-out> e.g.: gmm-acc-stats-ali 1.mdl scp:train.scp ark:1.ali 1.acc |
| gmm-align.cc | Align features given [GMM-based] models. Usage: gmm-align [options] tree-in model-in lexicon-fst-in feature-rspecifier transcriptions-rspecifier alignments-wspecifier e.g.: gmm-align tree 1.mdl lex.fst scp:train.scp ark:train.tra ark:1.ali |
| gmm-decode-faster.cc | Decode features using GMM-based model. Usage: gmm-decode-faster [options] model-in fst-in features-rspecifier words-wspecifier [alignments-wspecifier [lattice-wspecifier]] Note: lattices, if output, will just be linear sequences; use gmm-latgen-faster if you want "real" lattices. |
| gmm-decode-simple.cc | Decode features using GMM-based model. Usage: gmm-decode-simple [options] model-in fst-in features-rspecifier words-wspecifier [alignments-wspecifier] |
| gmm-decode-nbest.cc | Decode features using GMM-based model. Usage: gmm-decode-faster [options] model-in fst-in features-rspecifier nbestlattice-wspecifier words-wspecifier [alignments-wspecifier] |
| gmm-align-compiled.cc | Align features given [GMM-based] models. Usage: gmm-align-compiled [options] model-in graphs-rspecifier feature-rspecifier alignments-wspecifier e.g.: gmm-align-compiled 1.mdl ark:graphs.fsts scp:train.scp ark:1.ali or: compile-train-graphs tree 1.mdl lex.fst ark:train.tra b, ark:- | \ gmm-align-compiled 1.mdl ark:- scp:train.scp t, ark:1.ali |
| gmm-sum-accs.cc | Sum multiple accumulated stats files for GMM training. Usage: gmm-sum-accs [options] stats-out stats-in1 stats-in2 ... |
| gmm-init-trans.cc | Initialize transition model given topo, tree and GMM (used for format conversion from HTK) Usage: gmm-init-trans <topology-in> <gmm-in> <tree-in> <model-out> |
| gmm-est-regtree-fmllr.cc | Compute FMLLR transforms per-utterance (default) or per-speaker for the supplied set of speakers (spk2utt option). Note: writes RegtreeFmllrDiagGmm objects Usage: gmm-est-regtree-fmllr [options] <model-in> <feature-rspecifier> <posteriors-rspecifier> <regression-tree> <transforms-wspecifier> |
| gmm-acc-stats-twofeats.cc | Accumulate stats for GMM training, computing posteriors with one set of features but accumulating statistics with another. Usage: gmm-acc-stats-twofeats [options] <model-in> <feature1-rspecifier> <feature2-rspecifier> <posteriors-rspecifier> <stats-out> e.g.: gmm-acc-stats-twofeats 1.mdl 1.ali scp:train.scp scp:train_new.scp ark:1.ali 1.acc |
| gmm-acc-stats.cc | Accumulate stats for GMM training (reading in posteriors). Usage: gmm-acc-stats [options] <model-in> <feature-rspecifier><posteriors-rspecifier> <stats-out> e.g.: gmm-acc-stats 1.mdl scp:train.scp ark:1.post 1.acc |
| gmm-init-lvtln.cc | Initialize lvtln transforms Usage: gmm-init-lvtln [options] <lvtln-out> e.g.: gmm-init-lvtln --dim=13 --num-classes=21 --default-class=10 1.lvtln |
| gmm-est-lvtln-trans.cc | Estimate linear-VTLN transforms, either per utterance or for the supplied set of speakers (spk2utt option). Reads posteriors. Usage: gmm-est-lvtln-trans [options] <model-in> <lvtln-in> <feature-rspecifier> <gpost-rspecifier> <lvtln-trans-wspecifier> [<lvtln-class-wspecifier>] |
| gmm-train-lvtln-special.cc | Set one of the transforms in lvtln to the minimum-squared-error solution to mapping feats-untransformed to feats-transformed; alignments may optionally be used to downweight/remove silence. Usage: gmm-train-lvtln-special [options] class-index <lvtln-in> <lvtln-out> <feats-untransformed-rspecifier> <feats-transformed-rspecifier> [<posteriors-rspecifier>] e.g.: gmm-train-lvtln-special 5 5.lvtln 6.lvtln scp:train.scp scp:train_warp095.scp ark:nosil.post |
| gmm-acc-mllt.cc | Accumulate MLLT (global STC) statistics Usage: gmm-acc-mllt [options] <model-in> <feature-rspecifier> <posteriors-rspecifier> <stats-out> e.g.: gmm-acc-mllt 1.mdl scp:train.scp ark:1.post 1.macc |
| gmm-mixup.cc | Does GMM mixing up (and Gaussian merging) Usage: gmm-mixup [options] <model-in> <state-occs-in> <model-out> e.g. of mixing up: gmm-mixup --mix-up=4000 1.mdl 1.occs 2.mdl e.g. of merging: gmm-mixup --merge=2000 1.mdl 1.occs 2.mdl |
| gmm-init-model.cc | Initialize GMM from decision tree and tree stats Usage: gmm-init-model [options] <tree-in> <tree-stats-in> <topo-file> <model-out> [<old-tree> <old-model>] e.g.: gmm-init-model tree treeacc topo 1.mdl or (initializing GMMs with old model): gmm-init-model tree treeacc topo 1.mdl prev/tree prev/30.mdl |
| gmm-acc-hlda.cc | Accumulate HLDA statistics Usage: gmm-acc-hlda [options] <model-in> <orig-transform-in> <orig-feature-rspecifier> <posteriors-rspecifier> <stats-out> Note: orig-transform-in must be the current truncated HLDA transform (e.g. from LDA).e.g.: gmm-acc-hlda 1.mdl 1.hlda "ark:splice-feats scp:train.scp |" ark:1.post 1.hacc |
| gmm-est-hlda.cc | Do HLDA update Usage: gmm-est-hlda [options] <model-in> <full-hlda-mat-in> <model-out> <full-hlda-mat-out> <partial-hlda-mat-out> <stats-in1> <stats-in2> ... e.g.: gmm-est-hlda 1.mdl 1.hldafull 2.mdl 2.hldafull 2.hlda 1.0.hacc 1.1.hacc ... |
| gmm-transform-means.cc | Transform GMM means with linear or affine transform Usage: gmm-transform-means <transform-matrix> <model-in> <model-out> e.g.: gmm-transform-means 2.mat 2.mdl 3.mdl |
| gmm-init-et.cc | Initialize exponential tranform Usage: gmm-init-et [options] <et-object-out> e.g.: gmm-init-et --dim=39 1.et |
| gmm-est-et.cc | Compute exponential transforms (which are a special case of fMLLR transforms) per-utterance (default) or per-speaker for the supplied set of speakers (spk2utt option). Usage: gmm-est-et [options] <model> <exponential-transform> <feature-rspecifier> <gpost-rspecifier> <transforms-wspecifier> [<warp-factors-wspecifier>] |
| gmm-et-acc-a.cc | Accumulate statistics for estimating the A matrix of exponential transform, per-utterance (default) or per-speaker for the supplied set of speakers (spk2utt option). Usage: gmm-et-acc-a [options] <model> <exponential-transform> <feature-rspecifier> <gpost-rspecifier> <accs-filename> |
| gmm-et-est-a.cc | Update matrix A of exponential transform (uses stats from gmm-et-acc-a) Usage: gmm-et-est-a [options] <et-in> <et-out> <a-stats1> <a-stats2> ... e.g.: gmm-et-est-a 1.et 2.et 1.et_acc_a |
| gmm-copy-et.cc | Copy exponential transform object (possibly changing normalization type) Usage: gmm-copy-et [options] <et-object-in> <et-object-out> e.g.: gmm-copy-et --normalize-type=mean-and-var 1.et 2.et |
| gmm-et-get-b.cc | Write out the B matrix from the exponential transform (excluding last row), to a single file This can be treated as the "default" value of the exponential transform. Optionally, writes out the ET object with the B transform set to unity Usage: gmm-et-get-b [options] <et-object-in> <matrix-out> [<et-object-out>] e.g.: gmm-et-get-b 12.et B.mat final.et |
| gmm-et-get-a.cc | Write out the A matrix from the exponential transform to a single file Usage: gmm-et-get-a [options] <et-object-in> <matrix-out> e.g.: gmm-et-get-a --binary=false 12.et A.mat |
| gmm-make-regtree.cc | Build regression class tree. Usage: gmm-make-regtree [options] <model-file> <regtree-out> E.g.: gmm-make-regtree --silphones=1:2:3 --state-occs=1.occs 1.mdl 1.regtree [Note: state-occs come from --write-occs option of gmm-est] |
| gmm-decode-faster-regtree-fmllr.cc | Decode features using GMM-based model. Usage: gmm-decode-faster-regtree-fmllr [options] model-in fst-in regtree-in features-rspecifier transforms-rspecifier words-wspecifier [alignments-wspecifier] |
| gmm-post-to-gpost.cc | Convert state-level posteriors to Gaussian-level posteriors Usage: gmm-post-to-gpost [options] <model-in> <feature-rspecifier> <posteriors-rspecifier> <gpost-wspecifier> e.g.: gmm-post-to-gpost 1.mdl scp:train.scp ark:1.post ark:1.gpost |
| gmm-est-fmllr-gpost.cc | Estimate global fMLLR transforms, either per utterance or for the supplied set of speakers (spk2utt option). Reads Gaussian-level posteriors. Writes to a table of matrices. Usage: gmm-est-fmllr-gpost [options] <model-in> <feature-rspecifier> <gpost-rspecifier> <transform-wspecifier> |
| gmm-est-fmllr.cc | Estimate global fMLLR transforms, either per utterance or for the supplied set of speakers (spk2utt option). Reads Gaussian-level posteriors. Writes to a table of matrices. Usage: gmm-est-fmllr [options] <model-in> <feature-rspecifier> <post-rspecifier> <transform-wspecifier> |
| gmm-est-regtree-fmllr-ali.cc | Compute FMLLR transforms per-utterance (default) or per-speaker for the supplied set of speakers (spk2utt option). Note: writes RegtreeFmllrDiagGmm objects Usage: gmm-est-regtree-fmllr-ali [options] <model-in> <feature-rspecifier> <alignments-rspecifier> <regression-tree> <transforms-wspecifier> |
| gmm-est-regtree-mllr.cc | Compute MLLR transforms per-utterance (default) or per-speaker for the supplied set of speakers (spk2utt option). Note: writes RegtreeMllrDiagGmm objects Usage: gmm-est-regtree-mllr [options] <model-in> <feature-rspecifier> <posteriors-rspecifier> <regression-tree> <transforms-wspecifier> |
| gmm-decode-kaldi.cc | Decode features using GMM-based model. Usage: gmm-decode-kaldi [options] model-in fst-in features-rspecifier words-wspecifier [lattice-wspecifier] |
| gmm-compute-likes.cc | Compute log-likelihoods from GMM-based model (outputs matrices of log-likelihoods indexed by (frame, pdf) Usage: gmm-compute-likes [options] model-in features-rspecifier likes-wspecifier |
| gmm-decode-faster-regtree-mllr.cc | Decode features using GMM-based model. Usage: gmm-decode-faster-regtree-mllr [options] model-in fst-in regtree-in features-rspecifier transforms-rspecifier words-wspecifier [alignments-wspecifier] |
| gmm-et-apply-c.cc | Given the matrix Cpart which represents an MLLT/STC transform, let C = Cpart extended with zeros (and one as the new diagonal element); update the A and B matrices with A := C A C^{-1} and B := C B Usage: gmm-et-apply-c [options] <et-object-in> <c-matrix-in> <et-object-out> e.g.: gmm-et-apply-c 1.et C.mat 2.et |
| gmm-latgen-simple.cc | Generate lattices using GMM-based model. Usage: gmm-latgen-simple [options] model-in fst-in features-rspecifier lattice-wspecifier [ words-wspecifier [alignments-wspecifier] ] |
| gmm-rescore-lattice.cc | Replace the acoustic scores on a lattice using a new model. Usage: gmm-rescore-lattice [options] <model-in> <lattice-rspecifier> <feature-rspecifier> <lattice-wspecifier> e.g.: gmm-rescore-lattice 1.mdl ark:1.lats scp:trn.scp ark:2.lats |
| gmm-decode-biglm-faster.cc | Decode features using GMM-based model. User supplies LM used to generate decoding graph, and desired LM; this decoder applies the difference during decoding Usage: gmm-decode-biglm-faster [options] model-in fst-in oldlm-fst-in newlm-fst-in features-rspecifier words-wspecifier [alignments-wspecifier [lattice-wspecifier]] |
| gmm-est-gaussians-ebw.cc | Do EBW update for MMI, MPE or MCE discriminative training. Numerator stats should already be I-smoothed (e.g. use gmm-ismooth-stats) Usage: gmm-est-gaussians-ebw [options] <model-in> <stats-num-in> <stats-den-in> <model-out> e.g.: gmm-est-gaussians-ebw 1.mdl num.acc den.acc 2.mdl |
| gmm-est-weights-ebw.cc | Do EBW update on weights for MMI, MPE or MCE discriminative training. Numerator stats should not be I-smoothed Usage: gmm-est-weights-ebw [options] <model-in> <stats-num-in> <stats-den-in> <model-out> e.g.: gmm-est-weights-ebw 1.mdl num.acc den.acc 2.mdl |
| gmm-latgen-faster.cc | Generate lattices using GMM-based model. Usage: gmm-latgen-faster [options] model-in (fst-in|fsts-rspecifier) features-rspecifier lattice-wspecifier [ words-wspecifier [alignments-wspecifier] ] |
| gmm-copy.cc | Copy GMM based model (and possibly change binary/text format) Usage: gmm-copy [options] <model-in> <model-out> e.g.: gmm-copy --binary=false 1.mdl 1_txt.mdl |
| gmm-global-acc-stats.cc | Accumulate stats for training a diagonal-covariance GMM. Usage: gmm-global-acc-stats [options] <model-in> <feature-rspecifier> <stats-out> e.g.: gmm-global-acc-stats 1.mdl scp:train.scp 1.acc |
| gmm-global-est.cc | Estimate a diagonal-covariance GMM from the accumulated stats. Usage: gmm-global-est [options] <model-in> <stats-in> <model-out> |
| gmm-global-sum-accs.cc | Sum multiple accumulated stats files for diagonal-covariance GMM training. Usage: gmm-global-sum-accs [options] stats-out stats-in1 stats-in2 ... |
| gmm-gselect.cc | Precompute Gaussian indices for pruning (e.g. in training UBMs, SGMMs, tied-mixture systems) For each frame, gives a list of the n best Gaussian indices, sorted from best to worst. Usage: gmm-gselect [options] <model-in> <feature-rspecifier> <gselect-wspecifier> The --gselect option (which takes an rspecifier) limits selection to a subset of indices: e.g.: gmm-gselect "--gselect=ark:gunzip -c bigger.gselect.gz|" --n=20 1.gmm "ark:feature-command |" "ark,t:|gzip -c >1.gselect.gz" |
| gmm-latgen-biglm-faster.cc | Generate lattices using GMM-based model. User supplies LM used to generate decoding graph, and desired LM; this decoder applies the difference during decoding Usage: gmm-latgen-biglm-faster [options] model-in (fst-in|fsts-rspecifier) oldlm-fst-in newlm-fst-in features-rspecifier lattice-wspecifier [ words-wspecifier [alignments-wspecifier] ] |
| gmm-ismooth-stats.cc | Apply I-smoothing to statistics, e.g. for discriminative training Usage: gmm-ismooth-stats [options] [--smooth-from-model] [<src-stats-in>|<src-model-in>] <dst-stats-in> <stats-out> e.g.: gmm-ismooth-stats --tau=100 ml.acc num.acc smoothed.acc or: gmm-ismooth-stats --tau=50 --smooth-from-model 1.mdl num.acc smoothed.acc or: gmm-ismooth-stats --tau=100 num.acc num.acc smoothed.acc |
| gmm-global-get-frame-likes.cc | Print out per-frame log-likelihoods for each utterance, as an archive of vectors of floats. Usage: gmm-global-get-frame-likes [options] <model-in> <feature-rspecifier> <likes-out-wspecifier> e.g.: gmm-global-get-frame-likes 1.mdl scp:train.scp ark:1.likes |
| gmm-global-est-fmllr.cc | Estimate global fMLLR transforms, either per utterance or for the supplied set of speakers (spk2utt option). Reads features, and (with --weights option) weights for each frame (also see --gselect option) Usage: gmm-global-est-fmllr [options] <gmm-in> <feature-rspecifier> <transform-wspecifier> |
| gmm-global-to-fgmm.cc | Convert single diagonal-covariance GMM to single full-covariance GMM. Usage: gmm-global-to-fgmm [options] 1.gmm 1.fgmm |
| gmm-global-acc-stats-twofeats.cc | Accumulate stats for training a diagonal-covariance GMM, two-feature version Usage: gmm-global-acc-stats-twofeats [options] <model-in> <feature1-rspecifier> <feature2-rspecifier> <stats-out> e.g.: gmm-global-acc-stats-twofeats 1.mdl scp:train.scp scp:train2.scp 1.acc |
| gmm-global-copy.cc | Copy a diagonal-covariance GMM Usage: gmm-global-copy [options] <model-in> <model-out> e.g.: gmm-global-copy --binary=false 1.model - | less |
| gmm-align-compiled-plusphones.cc | Align features given [GMM-based] models, but adds in likelihoods of simple per-phone GMMs with alpha*per-phone-like + (1-alpha)*model-like. This gives more consistent alignments. Per-phone models are obtained by K-means on weighted model states, using the transition-accs to get weights. (e.g. use the first line of text format of normal accs). Note: this program actually isn't that useful. We keep it mainly as an example of how to write a decoder with interpolated likelihoods. Usage: gmm-align-compiled-plusphones [options] transition-accs-in model-in graphs-rspecifier feature-rspecifier alignments-wspecifier e.g.: gmm-align-compiled-plusphones --alpha=0.2 --acoustic-scale=0.1 \ 1.acc 1.mdl ark:graphs.fsts scp:train.scp ark:1.ali or: compile-train-graphs tree 1.mdl lex.fst ark:train.tra b, ark:- | \ gmm-align-compiled-plusphones 1.acc 1.mdl ark:- scp:train.scp t, ark:1.ali |
| gmm-get-feat-deriv.cc | From GMM model and posteriors (which don't have to be positive), output for each utterance a matrix of likelihood derivatives w.r.t. the features. E.g. used in feature-space discriminative training. Usage: gmm-get-feat-deriv [options] <model-in> <feature-rspecifier> <posteriors-rspecifier> <feature-deriv-wspecifier> e.g.: gmm-get-feat-deriv 1.mdl "$feats" ark:1.post ark:1.deriv |
| gmm-fmpe-acc-stats.cc | Accumulate stats for fMPE training, using GMM model. Note: this could be done using gmm-get-feat-deriv and fmpe-acc-stats (but you'd be computing the features twice). Features input should be pre-fMPE features. Usage: gmm-fmpe-acc-stats [options] <model-in> <fmpe-in> <feature-rspecifier> <gselect-rspecifier> <posteriors-rspecifier> <fmpe-stats-out> e.g.: gmm-fmpe-acc-stats 1.mdl 1.fmpe "$feats" ark:1.gselect ark:1.post 1.fmpe_stats |
| gmm-acc-stats2.cc | Accumulate stats for GMM training (from posteriors) This version writes two accumulators (e.g. num and den), and puts the positive accumulators in num, negative in den Usage: gmm-acc-stats2 [options] <model> <feature-rspecifier><posteriors-rspecifier> <num-stats-out> <den-stats-out> e.g.: gmm-acc-stats 1.mdl "$feats" ark:1.post 1.num_acc 1.den_acc |
| gmm-init-model-flat.cc | Initialize GMM, with Gaussians initialized to mean and variance of some provided example data Usage: gmm-init-model-flat [options] <tree-in> <topo-file> <model-out> <features-rspecifier> e.g.: gmm-init-model-flat tree topo 1.mdl ark:feats.scp |
| gmm-info.cc | Write to standard output various properties of GMM-based model Usage: gmm-info [options] <model-in> e.g.: gmm-info 1.mdl |
| lattice-best-path.cc | Generate 1-best path through lattices; output as transcriptions and alignments Note: if you want output as FSTs, use lattice-1best; if you want output with acoustic and LM scores, use lattice-1best | nbest-to-linear Usage: lattice-best-path [options] lattice-rspecifier [ transcriptions-wspecifier [ alignments-wspecifier] ] e.g.: lattice-best-path --acoustic-scale=0.1 ark:1.lats ark:1.tra ark:1.ali |
| lattice-prune.cc | Apply beam pruning to lattices Usage: lattice-prune [options] lattice-rspecifier lattice-wspecifier e.g.: lattice-prune --acoustic-scale=0.1 --beam=4.0 ark:1.lats ark:pruned.lats |
| lattice-equivalent.cc | Test whether sets of lattices are equivalent (return with status 0 if all were equivalent, 1 otherwise, -1 on error) Usage: lattice-equivalent [options] lattice-rspecifier1 lattice-rspecifier2 e.g.: lattice-equivalent ark:1.lats ark:2.lats |
| lattice-to-nbest.cc | Work out N-best paths in lattices and write out as FSTs Usage: lattice-to-nbest [options] lattice-rspecifier lattice-wspecifier e.g.: lattice-to-nbest --acoustic-scale=0.1 --n=10 ark:1.lats ark:nbest.lats |
| lattice-lmrescore.cc | Add lm_scale * [cost of best path through LM FST] to graph-cost of paths through lattice. Does this by composing with LM FST, then lattice-determinizing (it has to negate weights first if lm_scale<0) Usage: lattice-lmrescore [options] lattice-rspecifier lm-fst-in lattice-wspecifier e.g.: lattice-lmrescore --lm-scale=-1.0 ark:in.lats data/G.fst ark:out.lats |
| lattice-scale.cc | Apply scaling to lattice weights Usage: lattice-scale [options] lattice-rspecifier lattice-wspecifier e.g.: lattice-scale --lm-scale=0.0 ark:1.lats ark:scaled.lats |
| lattice-union.cc | Takes two archives of lattices (indexed by utterances) and computes the union of the individual lattice pairs (one from each archive). Usage: lattice-union [options] lattice-rspecifier1 lattice-rspecifier2 lattice-wspecifier e.g.: lattice-union ark:den.lats ark:num.lats ark:union.lats |
| lattice-to-post.cc | Do forward-backward and collect posteriors over lattices. Usage: lattice-to-post [options] lats-rspecifier posts-wspecifier [loglikes-wspecifier] e.g.: lattice-to-post --acoustic-scale=0.1 ark:1.lats ark:1.post |
| lattice-determinize.cc | lattice-determinize lattices (and apply a pruning beam) (see http://kaldi.sourceforge.net/lattices.html for more explanation) note: this program is tyically only useful if you generated state-level lattices, e.g. called gmm-latgen-simple with --determinize=false Usage: lattice-determinize [options] lattice-rspecifier lattice-wspecifier e.g.: lattice-determinize --acoustic-scale=0.1 --beam=15.0 ark:1.lats ark:det.lats |
| lattice-oracle.cc | Finds the path having the smallest edit-distance between two lattices. For efficiency put the smallest lattices first (for example reference strings). Usage: lattice-oracle [options] test-lattice-rspecifier reference-rspecifier transcriptions-wspecifier e.g.: lattice-oracle ark:ref.lats ark:1.tra ark:2.tra |
| lattice-rmali.cc | Remove state-sequences from lattice weights Usage: lattice-rmali [options] lattice-rspecifier lattice-wspecifier e.g.: lattice-rmali ark:1.lats ark:proj.lats |
| lattice-compose.cc | Composes lattices (in transducer form, as type Lattice). Depending on the command-line arguments, either composes lattices with lattices, or lattices with FSTs (rspecifiers are assumed to be lattices, and rxfilenames are assumed to be FSTs, which have their weights interpreted as "graph weights" when converted into the Lattice format. Usage: lattice-compose [options] lattice-rspecifier1 (lattice-rspecifier2|fst-rxfilename2) lattice-wspecifier e.g.: lattice-compose ark:1.lats ark:2.lats ark:composed.lats or: lattice-compose ark:1.lats G.fst ark:composed.lats |
| lattice-boost-ali.cc | Boost graph likelihoods (decrease graph costs) by b * #frame-phone-errors on each arc in the lattice. Useful for discriminative training, e.g. boosted MMI. Modifies input lattices. This version takes the reference in the form of alignments. Needs the model (just the transitions) to transform pdf-ids to phones. Takes the --silence-phones option and these phones appearing in the lattice are always assigned zero error, or with the --max-silence-error option, at most this error-count per frame (--max-silence-error=1 is equivalent to not specifying --silence-phones). Usage: lattice-boost-ali [options] model lats-rspecifier ali-rspecifier lats-wspecifier e.g.: lattice-boost-ali --silence-phones=1:2:3 --b=0.05 1.mdl ark:1.lats ark:1.ali ark:boosted.lats |
| lattice-copy.cc | Copy lattices (e.g. useful for changing to text mode or changing format to standard from compact lattice.) Usage: lattice-copy [options] lattice-rspecifier lattice-wspecifier e.g.: lattice-copy --write-compact=false ark:1.lats ark,t:text.lats |
| lattice-to-fst.cc | Turn lattices into normal FSTs, retaining only the word labels By default, removes all weights and also epsilons (configure with with --acoustic-scale, --lm-scale and --rm-eps) Usage: lattice-to-fst [options] lattice-rspecifier fsts-wspecifier e.g.: lattice-to-fst ark:1.lats ark:1.fsts |
| lattice-to-phone-lattice.cc | Convert the words (or whatever other output labels are on the lattice) into phones, worked out from the alignment (needs the transition model to work out the phone sequence from the transition-ids) Usage: lattice-to-phone-lattice [options] model lattice-rspecifier lattice-wspecifier e.g.: lattice-to-phone-lattice 1.mdl ark:1.lats ark:phones.lats |
| lattice-interp.cc | Takes two archives of lattices (indexed by utterances) and combines the individual lattice pairs (one from each archive). Keeps the alignments from the first lattice. Equivalent to projecting the second archive on words (lattice-project), then composing the pairs of lattices (lattice-compose), then scaling graph and acoustic costs by 0.5 (lattice-scale). You can control the individual scales with --alpha, which is the scale of the first lattices (the second is 1-alpha). Usage: lattice-interp [options] lattice-rspecifier-a lattice-rspecifier-b lattice-wspecifier e.g.: lattice-compose ark:1.lats ark:2.lats ark:composed.lats |
| lattice-project.cc | Project lattices (in their transducer form); by default makes them word->word transducers (set --project-output=false for tid->tid). Usage: lattice-project [options] lattice-rspecifier lattice-wspecifier e.g.: lattice-project ark:1.lats ark:word2word.lats or: lattice-project --project-output=false ark:1.lats ark:tid2tid.lats |
| lattice-add-trans-probs.cc | Add transition probabilities into graph part of lattice scores, controlled by options --transition-scale and --self-loop-scale, which for compatibility with the original graph, would normally be set to the same values used in graph compilatoin Usage: lattice-add-trans-probs [options] model lattice-rspecifier lattice-wspecifier e.g.: lattice-add-trans-probs --transition-scale=1.0 --self-loop-scale=0.1 1.mdl ark:in.lats ark:out.lats |
| lattice-difference.cc | Compute FST difference on lattices (remove sequences in first lattice that appear in second lattice) Useful for the denominator lattice for MCE. Usage: lattice-difference [options] lattice1-rspecifier lattice2-rspecifier lattice-wspecifier e.g.: lattice-difference ark:den.lats ark:num.lats ark:den_mce.lats |
| lattice-word-align.cc | (note: from the s5 scripts onward, this is deprecated, see lattice-align-words) Create word-aligned lattices (in which the arcs correspond with word boundaries) Usage: lattice-word-align [options] <model> <lattice-rspecifier> <lattice-wspecifier> e.g.: lattice-word-align --silence-phones=1:2 --wbegin-phones=2:6:10:14 \ --wend-phones=3:7:11:15 --winternal-phones=4:8:12:16 --wbegin-and-end-phones=5:9:13:17 \ --silence-label=2 --partial-word-label=16342 \ final.mdl ark:1.lats ark:aligned.lats |
| nbest-to-linear.cc | Takes as input lattices/n-bests which must be linear (single path); convert from lattice to up to 4 archives containing transcriptions, alignments, and acoustic and LM costs (note: use ark:/dev/null for unwanted outputs) Usage: nbest-to-linear [options] <nbest-rspecifier> <transcriptions-wspecifier> [<alignments-wspecifier> [<lm-cost-wspecifier> [<ac-cost-wspecifier>]]] e.g.: lattice-to-nbest --n=10 ark:1.lats ark:- | \ nbest-to-linear ark:1.lats ark,t:1.ali ark,t:1.tra |
| nbest-to-lattice.cc | Read in a Table containing N-best entries from a lattices (i.e. individual lattices with a linear structure, one for each N-best entry, indexed by utt_id_a-1, utt_id_a-2, etc., and take the union of them for each utterance id (e.g. utt_id_a), outputting a lattice for each. Usage: nbest-to-lattice <nbest-rspecifier> <lattices-wspecifier> e.g.: nbest-to-lattice ark:1.nbest ark:1.lats |
| latbin/lattice-1best.cc "lattice-1best.cc" | Compute best path through latties and write out as FSTs Note: differs from lattice-nbest with --n=1 because we won't append -1 to the utterance-ids. Differs from lattice-best-path because output is FST. Usage: lattice-1best [options] lattice-rspecifier lattice-wspecifier e.g.: lattice-1best --acoustic-scale=0.1 ark:1.lats ark:1best.lats |
| linear-to-nbest.cc | This does the opposite of nbest-to-linear. It takes 4 archives, containing alignments, word-sequences, and acoustic and LM costs, and turns it into a single archive containing FSTs with a linear structure. The program is called linear-to-nbest because very often the archives concerned will represent N-best lists Usage: linear-to-nbest [options] <alignments-rspecifier> <transcriptions-rspecifier> <lm-cost-rspecifier> <ac-cost-rspecifier> <nbest-wspecifier> e.g.: linear-to-nbest ark:1.ali ark:1.tra ark:1.lmscore ark:1.acscore ark:1.nbest |
| lattice-mbr-decode.cc | Do Minimum Bayes Risk decoding (decoding that aims to minimize the expected word error rate). Possible outputs include the 1-best path (i.e. the word-sequence, as a sequence of ints per utterance), the computed Bayes Risk for each utterance, and the sausage stats as (for each utterance) std::vector<std::vector<std::pair<int32, float> > > for which we use the same I/O routines as for posteriors (type Posterior). times-wspecifier writes pairs of (start-time, end-time) in frames, for each sausage position, or for each one-best entry if --one-best-times=true. Note: use ark:/dev/null or the empty string for unwanted outputs. Note: times will only be very meaningful if you first use lattice-word-align. Usage: lattice-mbr-decode [options] lattice-rspecifier transcriptions-wspecifier [ bayes-risk-wspecifier [ sausage-stats-wspecifier [ times-wspecifier] ] ] e.g.: lattice-mbr-decode --acoustic-scale=0.1 ark:1.lats ark:1.tra ark:/dev/null ark:1.sau |
| lattice-align-words.cc | Convert lattices so that the arcs in the CompactLattice format correspond with words (i.e. aligned with word boundaries). Usage: lattice-align-words [options] <word-boundary-file> <model> <lattice-rspecifier> <lattice-wspecifier> e.g.: lattice-align-words --silence-label=4320 --partial-word-label=4324 \ data/lang/phones/word_boundary.int final.mdl ark:1.lats ark:aligned.lats Note: word-boundary file has format (on each line): <integer-phone-id> [begin|end|singleton|internal|nonword] |
| init-ubm.cc | Cluster the Gaussians in a diagonal-GMM acoustic model to a single full-covariance or diagonal-covariance GMM. Usage: init-ubm [options] <model-file> <state-occs> <gmm-out> |
| sgmm-align-compiled.cc | Align features given [SGMM-based] models. Usage: sgmm-align-compiled [options] model-in graphs-rspecifier feature-rspecifier alignments-wspecifier e.g.: sgmm-align-compiled 1.mdl ark:graphs.fsts scp:train.scp ark:1.ali |
| sgmm-acc-stats-ali.cc | Accumulate stats for SGMM training. Usage: sgmm-acc-stats-ali [options] <model-in> <feature-rspecifier> <alignments-rspecifier> <stats-out> e.g.: sgmm-acc-stats-ali 1.mdl 1.ali scp:train.scp ark:1.ali 1.acc |
| sgmm-sum-accs.cc | Sum multiple accumulated stats files for SGMM training. Usage: sgmm-sum-accs [options] stats-out stats-in1 stats-in2 ... |
| sgmm-est.cc | Estimate SGMM model parameters from accumulated stats. Usage: sgmm-est [options] <model-in> <stats-in> <model-out> |
| sgmm-decode-faster.cc | Decode features using SGMM-based model. Usage: sgmm-decode-faster [options] <model-in> <fst-in> <features-rspecifier> <words-wspecifier> [alignments-wspecifier] |
| sgmm-init.cc | Initialize an SGMM from a trained full-covariance UBM and a specified model topology. Usage: sgmm-init [options] <topology> <tree> <init-model> <sgmm-out> The <init-model> argument can be a UBM (the default case) or another SGMM (if the --init-from-sgmm flag is used). |
| sgmm-gselect.cc | Precompute Gaussian indices for SGMM training Usage: sgmm-gselect [options] <model-in> <feature-rspecifier> <gselect-wspecifier> e.g.: sgmm-gselect 1.sgmm "ark:feature-command |" ark:1.gs Note: you can do the same thing by combining the programs sgmm-write-ubm, fgmm-global-to-gmm, gmm-gselect and fgmm-gselect |
| sgmm-est-fmllr.cc | Estimate FMLLR transform for SGMMs, either per utterance or for the supplied set of speakers (with spk2utt option). Reads state-level posteriors. Writes to a table of matrices. Usage: sgmm-est-fmllr [options] <model-in> <feature-rspecifier> <post-rspecifier> <mats-wspecifier> |
| sgmm-acc-stats.cc | Accumulate stats for SGMM training. Usage: sgmm-acc-stats [options] <model-in> <feature-rspecifier> <posteriors-rspecifier> <stats-out> e.g.: sgmm-acc-stats 1.mdl 1.ali scp:train.scp 'ark:ali-to-post 1.ali ark:-|' 1.acc |
| sgmm-est-spkvecs.cc | Estimate SGMM speaker vectors, either per utterance or for the supplied set of speakers (with spk2utt option). Reads Gaussian-level posteriors. Writes to a table of vectors. Usage: sgmm-est-spkvecs [options] <model-in> <feature-rspecifier> <post-rspecifier> <vecs-wspecifier> |
| sgmm-post-to-gpost.cc | Convert posteriors to Gaussian-level posteriors for SGMM training. Usage: sgmm-post-to-gpost [options] <model-in> <feature-rspecifier> <posteriors-rspecifier> <gpost-wspecifier> e.g.: sgmm-post-to-gpost 1.mdl 1.ali scp:train.scp 'ark:ali-to-post ark:1.ali ark:-|' ark:- |
| sgmm-acc-stats-gpost.cc | Accumulate stats for SGMM training, given Gaussian-level posteriors Usage: sgmm-acc-stats-gpost [options] <model-in> <feature-rspecifier> <gpost-rspecifier> <stats-out> e.g.: sgmm-acc-stats-gpost 1.mdl 1.ali scp:train.scp ark, s, cs:- 1.acc |
| sgmm-est-spkvecs-gpost.cc | Estimate SGMM speaker vectors, either per utterance or for the supplied set of speakers (with spk2utt option). Reads Gaussian-level posteriors. Writes to a table of vectors. Usage: sgmm-est-spkvecs-gpost [options] <model-in> <feature-rspecifier> <gpost-rspecifier> <vecs-wspecifier> |
| sgmm-comp-prexform.cc | Compute "pre-transform" parameters required for estimating fMLLR with SGMMs, and write to a model file, after the SGMM. Usage: sgmm-comp-prexform [options] <sgmm-in> <occs-in> <sgmm-out> |
| sgmm-est-fmllr-gpost.cc | Estimate FMLLR transform for SGMMs, either per utterance or for the supplied set of speakers (with spk2utt option). Reads Gaussian-level posteriors. Writes to a table of matrices. Usage: sgmm-est-fmllr-gpost [options] <model-in> <feature-rspecifier> <gpost-rspecifier> <mats-wspecifier> |
| sgmm-acc-fmllrbasis-ali.cc | Accumulate stats for FMLLR bases training. Usage: sgmm-acc-fmllrbasis-ali [options] <model-in> <feature-rspecifier> <alignments-rspecifier> <spk2utt-rspecifier> <stats-out> e.g.: sgmm-acc-fmllrbasis-ali 1.mdl scp:train.scp ark:1.ali 1.acc |
| sgmm-est-fmllrbasis.cc | Sum multiple accumulated stats files for SGMM training. Usage: sgmm-est-fmllrbasis [options] <model-in> <model-out> <stats-in1> [stats-in2 ...] |
| sgmm-calc-distances.cc | Compute matrix of approximated K-L divergences between states Only works properly if a single substate per state. Usage: sgmm-calc-distances [options] model-in occs-in distances-out |
| sgmm-normalize.cc | Renormalize SGMM so that within certain subsets of UBM Gaussians (typically corresponding to gender), probabilities sum to one; write it out, including normalizers.Note: gaussians-rspecifier will normally be "ark:foo" where foo looks like m 0 1 2 3 4 5 f 6 7 8 9 10 Usage: sgmm-normalize [options] <model-in> <gaussians-rspecifier> <model-out> |
| sgmm-latgen-simple.cc | Decode features using SGMM-based model. Usage: sgmm-latgen-simple [options] <model-in> <fst-in> <features-rspecifier> <lattices-wspecifier> [<words-wspecifier> [<alignments-wspecifier>] ] |
| sgmm-latgen-faster.cc | Decode features using SGMM-based model. Usage: sgmm-latgen-faster [options] <model-in> (<fst-in>|<fsts-rspecifier>) <features-rspecifier> <lattices-wspecifier> [<words-wspecifier> [<alignments-wspecifier>] ] |
| sgmm-rescore-lattice.cc | Replace the acoustic scores on a lattice using a new model. Usage: sgmm-rescore-lattice [options] <model-in> <lattice-rspecifier> <feature-rspecifier> <lattice-wspecifier> e.g.: sgmm-rescore-lattice 1.mdl ark:1.lats scp:trn.scp ark:2.lats |
| sgmm-copy.cc | Copy SGMM (possibly changing binary/text format) Usage: sgmm-normalize [options] <model-in> <model-out> e.g.: sgmm-normalize --binary=false 1.mdl 1_text.mdl |
| sgmm-write-ubm.cc | Write out the full-covariance UBM of the SGMM Usage: sgmm-write-ubm [options] <model-in> <ubm-out> e.g.: sgmm-write-ubm 1.mdl 1.ubm |
| sgmm-mixup.cc | Increase number of sub-states or dimensions in SGMM Usage: sgmm-mixup [options] <model-in> <model-out> E.g. of mixing up: sgmm-mixup --read-occs=1.occs --num-substates=10000 1.mdl 2.mdl E.g. of increasing phonetic dim: sgmm-mixup --increase-phn-dim=50 1.mdl 2.mdl E.g. of increasing speaker dim: sgmm-mixup --increase-spk-dim=50 1.mdl 2.mdl E.g. of removing speaker space: sgmm-mixup --remove-speaker-space 1.mdl 2.mdl These modes may be combined. |
| sgmm-info.cc | Print various information about an SGMM. Usage: sgmm-info [options] <model-in> |
| sgmm-acc-tree-stats.cc | Accumulate statistics for decision tree training. This version accumulates statistics in the form of state-specific SGMM stats; you need to use the program sgmm-build-tree to build the tree (and sgmm-sum-tree-accs to sum the stats). Usage: sgmm-acc-tree-stats [options] sgmm-model-in features-rspecifier alignments-rspecifier [tree-accs-out] e.g.: sgmm-acc-tree-stats --ci-phones=48:49 1.mdl scp:train.scp ark:1.ali 1.tacc |
| sgmm-sum-tree-stats.cc | Sum SGMM-type statistics used for phonetic decision tree building. Usage: sgmm-sum-tree-stats [options] tree-accs-out trea-accs-in1 tree-accs-in2 ... e.g.: sgmm-sum-tree-stats treeacc 1.streeacc 2.streeacc 3.streeacc |
| sgmm-build-tree.cc | Train decision tree Usage: sgmm-build-tree [options] <old-sgmm-in> <tree-stats-in> <roots-file> <questions-file> <tree-out> [<sgmm-out>] e.g.: sgmm-build-tree 0.sgmm streeacc roots.txt 1.qst tree |
| sgmm-cluster-phones.cc | Cluster phones (or sets of phones) into sets for various purposes Usage: sgmm-cluster-phones [options] <sgmm-in> <tree-stats-in> <phone-sets-in> <clustered-phones-out> e.g.: sgmm-cluster-phones 0.sgmm 1.tacc phonesets.txt questions.txt |
| sgmm-init-from-tree-stats.cc | Initialize an SGMM from a previously built SGMM, a tree, and SGMM-type tree stats Usage: sgmm-init-from-tree-stats [options] <old-sgmm> <tree> <sgmm-tree-stats> <sgmm-out> |
| tied-diag-gmm-init-mono.cc | Initialize monophone GMM with tied mixtures. Usage: tied-diag-gmm-init-mono <topology-in> <diag-codebook> <model-out> <tree-out> e.g.: tied-diag-gmm-init-mono topo cb.pdf mono.mdl mono.tree |
| tied-diag-gmm-acc-stats-ali.cc | Accumulate stats for tied GMM training. Usage: tied-diag-gmm-acc-stats-ali [options] <model-in> <feature-rspecifier> <alignments-rspecifier> <stats-out> e.g.: tied-diag-gmm-acc-stats-ali 1.mdl 1.ali scp:train.scp ark:1.ali 1.acc |
| tied-diag-gmm-est.cc | Accumulate stats for GMM training. Usage: tied-diag-gmm-est [options] <model-in> <stats-in> <model-out> e.g.: tied-diag-gmm-est 1.mdl 1.acc 2.mdl |
| tied-diag-gmm-align-compiled.cc | Align features given tied diagonal GMM-based models. Usage: tied-diag-gmm-align-compiled [options] model-in graphs-rspecifier feature-rspecifier alignments-wspecifier e.g.: tied-diag-gmm-align-compiled 1.mdl ark:graphs.fsts scp:train.scp ark:1.ali or: compile-train-graphs tree 1.mdl lex.fst ark:train.tra b, ark:- | \ tied-diag-gmm-align-compiled 1.mdl ark:- scp:train.scp t, ark:1.ali |
| tied-diag-gmm-decode-simple.cc | Decode features using tied diagonal GMM-based model. Usage: tied-diag-gmm-decode-simple [options] model-in fst-in features-rspecifier words-wspecifier [alignments-wspecifier] |
| tied-diag-gmm-decode-faster.cc | Decode features using GMM-based model. Usage: gmm-decode-faster [options] model-in fst-in features-rspecifier words-wspecifier [alignments-wspecifier] |
| tied-diag-gmm-init-model.cc | Initialize a tied mixture model with diagonal mixture codebooks. If using more than one codebook, you need to specify a map file, mapping the tree leaves to the codebook ids as a vector, e.g. "[ 0 0 1 1 "] Usage: tied-diag-gmm-init-model [options] <tree> <topo> [<diag-ubm0> | <tree-map> <diag-ubm> ...] <model-out> e.g.: tied-diag-gmm-init-model tree topo tree.map diag0.ubm diag1.ubm 1.mdl |
| smooth-stats-diag.cc | Smooth the sufficient statistics of tied models by propagating them up and interpolating them down in the tree hierarchy (w.r.t. their codebook) Usage: smooth-stats-diag [options] <tree> <tree.map> <acc-in> <acc-out> e.g.: smooth-stats-diag tree tree.map 12.acc 12.acc.s |
| tied-full-gmm-init-mono.cc | Initialize monophone GMM with tied mixtures. Usage: tied-full-gmm-init-mono <topology-in> <full-codebook> <model-out> <tree-out> e.g.: tied-full-gmm-init-mono topo cb.pdf mono.mdl mono.tree |
| tied-full-gmm-acc-stats-ali.cc | Accumulate stats for tied GMM training. Usage: tied-full-gmm-acc-stats-ali [options] <model-in> <feature-rspecifier> <alignments-rspecifier> <stats-out> e.g.: tied-full-gmm-acc-stats-ali 1.mdl 1.ali scp:train.scp ark:1.ali 1.acc |
| tied-full-gmm-est.cc | Accumulate stats for GMM training. Usage: tied-full-gmm-est [options] <model-in> <stats-in> <model-out> e.g.: tied-full-gmm-est 1.mdl 1.acc 2.mdl |
| tied-full-gmm-align-compiled.cc | Align features given tied fullonal GMM-based models. Usage: tied-full-gmm-align-compiled [options] model-in graphs-rspecifier feature-rspecifier alignments-wspecifier e.g.: tied-full-gmm-align-compiled 1.mdl ark:graphs.fsts scp:train.scp ark:1.ali or: compile-train-graphs tree 1.mdl lex.fst ark:train.tra b, ark:- | \ tied-full-gmm-align-compiled 1.mdl ark:- scp:train.scp t, ark:1.ali |
| tied-full-gmm-decode-simple.cc | Decode features using tied fullonal GMM-based model. Usage: tied-full-gmm-decode-simple [options] model-in fst-in features-rspecifier words-wspecifier [alignments-wspecifier] |
| tied-full-gmm-decode-faster.cc | Decode features using GMM-based model. Usage: gmm-decode-faster [options] model-in fst-in features-rspecifier words-wspecifier [alignments-wspecifier] |
| tied-full-gmm-init-model.cc | Initialize a tied mixture model with fullonal mixture codebooks. If using more than one codebook, you need to specify a map file, mapping the tree leaves to the codebook ids as a vector, e.g. "[ 0 0 1 1 "] Usage: tied-full-gmm-init-model [options] <tree> <topo> [<diag-ubm0> | <tree-map> <diag-ubm> ...] <model-out> e.g.: tied-full-gmm-init-model tree topo tree.map full0.ubm full1.ubm 1.mdl |
| smooth-stats-full.cc | Smooth the sufficient statistics of tied models by propagating them up and interpolating them down in the tree hierarchy (w.r.t. their codebook) Usage: smooth-stats-full [options] <tree> <tree.map> <acc-in> <acc-out> e.g.: smooth-stats-full tree tree.map 12.acc 12.acc.s |
| init-tied-codebooks.cc | Generate codebooks for a tied mixture system based on the accumulated tree stats and optional two-level tree. Will write to <codebook-out>[.num] Usage: init-tied-codebooks [options] <tree-in> <tree-stats-in> <codebook-out> [tree.map] e.g.: init-tied-codebooks tree tree.acc ubm-full tree.map |
| tied-lbg.cc | Generate codebooks for a tied mixture system based on an existing tree and alignment, and write them to codebook-out[.num] If no tree.map is given, a single codebook generated. Usage: tied-lbg [options] tree-old tree-tied topo features-rspecifier alignments-rspecifier codebook-out [tree.map] e.g.: tied-lbg tree-old tree-tied topo scp:train.scp ark:ali ubm-full tree.map |
| full-to-diag.cc | Convert a full covariance GMM into a diagonal one. |
| Tool | Description |
1.7.3 
