  lib | |
   core | |
| filefmt | |
  bzip.c | Support for the bzip2 (bzlib) (de)compression methods |
| gzip.c | Support for the gzip (zlib) (de)compression methods |
| sphere.c | Support for the NIST SPHERE format |
 wav.c | Support for the RIFF-wav format |
| atomic.c | Provide a set of basic & advanced atomic operations |
| c_compat.c | Solve incompatibilities between the major platforms |
| datatype.c | Basic types |
| decodeargs.c | Interpretation of program arguments |
| dynmem.c | Dynamic memory management & checking |
| elementary_instr.c | Additional low-level functions |
| hash.c | Hash tables (and prime numbers) |
| ini_fini.c | Initialisation and termination |
| keyset.c | Key-sets |
| msg.c | Message and error processing |
| path_parse.c | Elementary path parsing |
| sig_handler.c | Signal handling |
| spr_oo.c | Provide the object oriented layer for the SPRAAK environment |
| srm.c | Shared Resource Management |
| str_parse.c | Elementary string parsing |
| stream.c | Unified data IO |
| utf8.c | Handling of UTF-8 encoded data |
| cwr | |
| am | |
| cwr_am_pp.c | Post-processing of acoustic log/linear probabilities/likelihoods |
| cwr_am_tbl.c | Collect all Acmod-modules |
| hmm_alloc.c | Allocation of HMM structures |
| hmm_ct_io.c | Read and write hmm training statistics |
| hmm_free.c | Free memory occupied by dynamically allocated HMM structures |
| hmm_print.c | All print-hmm-to-stream related subroutines |
| hmm_prog_util.c | Reusable high level operations on HMM's |
| hmm_read.c | All read-hmm-from-file related subroutines |
| hmm_setup.c | HMM initialising routines (when reading the HMM is done) |
| hmm_util.c | Some utility functions |
| hmm_vs_keys.c | Interference between HMM and data file keys |
| hmm_write.c | All write-hmm-to-file related subroutines |
| hmmdef.c | Definition of HMM constants |
| lm | |
| cwr_lm_cache.c | Extra layer on top of the LM-interface |
| cwr_lm_comb.c | |
| cwr_lm_dump.c | Dump all (non-cached) queries to the LM to a file |
| cwr_lm_ext.c | Extensions on standard LM's |
| cwr_lm_fsg.c | Finite state grammar |
| cwr_lm_interface.c | Define the LM-interface |
| cwr_lm_std.c | The standard N-gram LM |
| cwr_lm_tbl.c | Collect all LM-modules |
| cwr_lm_template.c | Template for a new LM |
| wlat | |
| wlat_bsearch.c | Beam search through the lattice |
| wlat_bsearch_.c | Beam search through the lattice |
| wlat_confidence.c | Add indicators to computer word confidence scores |
| wlat_master.c | Define the word lattice interface and provide the word lattice IO & hash operations |
| wlat_optimize.c | Optimize the word lattice |
| wlat_score.c | Rescore or compare word lattices |
| wlat_tbl.c | Collect all word lattice modules |
| check_mvg.c | Check and modify mvgs |
| ct_from_seg.c | Organisation of incremental count vectors |
| ct_to_oprob.c | Viterbi reestimation routines |
| cwr_compat1.c | Conversion between new storage formats and old storage formats |
| cwr_compat2.c | Convert command line options to the equivalent commands for cwr_main |
| cwr_compat3.c | Load the search space in old programs |
| cwr_count.c | Count events |
| cwr_err.c | Error rate for cwr-systems |
| cwr_fst.c | Finite state transducers – IO & conversion |
| cwr_lex.c | Handling of linear and tree structured lexicon descriptions |
| cwr_lmg.c | Handling of the language model |
| cwr_object_interface.c | A generic interface to load object for in the recognizer |
| cwr_phon_desc.c | Reading and conversion of phone descriptions + assimilation rules |
| cwr_search.c | Search routines for Large Vocabulary Continuous Speech Recognition |
| cwr_word.c | Handling of words in all cwr-routines |
| dft_am.c | Direct Feed Through (DFT) acoustic model |
| fast_mvg.c | Gaussian with diagonal covariance - fast evaluation |
| high_lvl_api.c | High level interface to the recognizer |
| lin_vs_log.c | Subroutines for mergestates and averwithref |
| mdt_am.c | Missing data techniques based acoustic modelling |
| mvg_eval.c | Computations of gaussian probabilities with diagonal covariance |
| output_prob.c | Calculates log of output probability of a dataframe for all states of a DD, an SC or a CD HMM |
| path2seg.c | Conversion between segmentations and path desriptions of alignments |
| phon_util.c | Handling of phoneme descriptions (homo phones) |
| pipe_sync.c | Synchronization between child and parent communicating via a pipe-pair |
| reject.c | Handling of rejection information |
| rm_gauss.c | Quick selection of gaussians for SC_HMMs |
| tree_util.c | Utilities for handling phonetic decision trees |
| data | |
| bssdef.c | Global variables |
| cor_seg.c | Hande corpus entries and segmentation descriptions |
| external_io.c | Bridge between the SPRAAK file format and external software such as Matlab |
| exemplar | |
| dtw_knn.c | Find the k nearest neighbour templates using DTW |
| exemplar_io.c | Read exemplar (meta) data |
| fvg | |
| fvg_ioe.c | Read and evaluate full variance gaussians |
| stats_2o.c | Calculate second order statistics on data |
| hash | |
| iptr_hash.c | Hash table for pointers and/or integers |
| str_hash.c | Hash string and other entries in a table |
| io | |
| line_ed.c | Line editing |
| legacy | |
| envvar.c | Environment variable handling |
| math | |
| erfinv.c | Inverse of the error function |
| fast_vec_math.c | Fast approximate vector functions |
| feval.c | Evaluation of expressions and handling of variables |
| fft.c | Good (fast, accurate, simple-to-use) FFT algorithm |
| lda.c | Calcultate the LDA transformation for a given set of data-points |
| math_basic.c | Basic matrix and vector operations |
| math_convert.c | Convert matrices between the different formats |
| math_eig.c | Solve the eigenvalue problem for a symmetric matrix |
| math_io.c | Write and read matrices in the new format |
| math_lu.c | LU-decomposition of matrices (symmetric and full) |
| matrix.c | Matrix input/output |
| random.c | Generating peudo random numbers |
| vector.c | Vector operations |
| mem | |
| memset.c | Initialize array of different types with a given constant value |
| mlp | |
| mlp_bwd_fun.c | Backward training of an MLP |
| mlp_eval.c | Fast evaluation of an MLP |
| mlp_eval_fun.c | Forward evaluation of an MLP |
| mlp_fwd_fun.c | Forward evaluation of an MLP |
| mlp_main.c | Main routines and types for handling MLP's |
| mlp_train.c | Training of an MLP |
| nn | |
| nn_functions.c | Define all linear and non-linear functions on the arcs |
| nn_io.c | Read/write the parameters for the linear and non-linear functions on the arcs |
| nn_main.c | Main routines and types for handling (deep) neural networks |
| preproc | |
| mod | |
| sspmod_add_noise.c | Add a controlled amount of noise to the incoming signal |
| sspmod_agc.c | Monitor the recording volume |
| sspmod_am_cvt.c | Conversion of acoustic likelihoods/probabilities |
| sspmod_asymm_filter.c | Asymmetric filtering |
| sspmod_aut2lpc.c | LPC from autocorrelation coefficients |
| sspmod_autocorr.c | Calculation of autocorrelation (and crosscorrelation) coefficients |
| sspmod_cor_info.c | Extract information from a corpus file |
| sspmod_deriv.c | Calculation of first and second time derivatives |
| sspmod_envelope.c | Calculate peak and dip envelope of a power spectrum |
| sspmod_equal_loud.c | Equal-loudness weighting |
| sspmod_feedback.c | Get feedback from the recognizer back into the preprocessing |
| sspmod_fft.c | Fast Fourier analysis and synthesis |
| sspmod_filter_bank.c | Design and simulate filter banks |
| sspmod_freq_warp.c | Transformation of a kid FFT power spectrum by including more formants |
| sspmod_fun_eval.c | Mathematical transformations of parameters |
| sspmod_histeq.c | Histogram based normalization of spectral features |
| sspmod_idct.c | Inverse Discrete Cosine Transform |
| sspmod_iir_filter.c | IIR filtering |
| sspmod_lin_comb.c | Linear combinations of the parameters |
| sspmod_lpc2cep.c | Convertion of lpc filter coefficients to cepstral coefficients |
| sspmod_math_fun.c | Execute some mathematical functions in a (ssp-)standard way, checking/updating keys if necessary |
| sspmod_matlab.c | Execution of a Matlab-routine in an SSP-script |
| sspmod_meannorm.c | Normalisation of parameters by subtraction of mean |
| sspmod_merge_fr.c | Merge a number of consecutive frames into one frame |
| sspmod_mlpeval.c | Evaluation of MLP's |
| sspmod_mv_svd_filt.c | Minimum Variance SVD-filtering of white noise in speech signals |
| sspmod_mvg.c | Evaluate a set of GMMs |
| sspmod_n_mean.c | Calculation of noise mean and variance |
| sspmod_nneval.c | Evaluation of Neural networks |
| sspmod_noise_masking.c | Normalisation of the SNR of a (mel) spectrum |
| sspmod_norm_aut.c | Normalisation of autocorrelation parameters |
| sspmod_online_bed.c | Begin/end point detection of speech |
| sspmod_ordstat.c | Ordered statistics |
| sspmod_pitch.c | Calculate pitch and voicing evidence |
| sspmod_sam2trk.c | Default processing of a sample file |
| sspmod_scale.c | Scaling of parameters |
| sspmod_snr_norm.c | Normalisation of dynamic range of filterbank energies |
| sspmod_spch_det.c | Speech Detection |
| sspmod_spec_impute.c | Impute a spectrum |
| sspmod_subsel_fr.c | Subselect frames |
| sspmod_time_filter.c | Filtering of samples |
| sspmod_trk2sam.c | Trk2sam turn overlapping frames with samples into a samplefile |
| sspmod_varfr.c | Variable frame rate analysis |
| sspmod_vq.c | Vector Quantization and Labeling |
| sspmod_vqmask.c | Impute a spectrum given a decomposed spectrum |
| sspmod_vtl_estim.c | Estimation of warping factors for vocal tract length (VTL) normalisation |
| cheby_design.c | Design of type II Chebychev filters and handling of filters in general |
| coding.c | Audio coding routines |
| conv_d.c | Polynomial convolution in direct form |
| fir_d.c | FIR filter implementation in direct form |
| fscale.c | Frequency transformations (perceptual) |
| fspec.c | Compute a spectrum from filter coefficients |
| ghfit.c | Computes the best n-gaussian fit to a histogram distribution |
| ghthr.c | Calculates n-1 thresholds in a n-gaussian probability distribution |
| iir_d.c | IIR filter implementation in direct form |
| lms.c | Calculates FIR-adaptive filters with LMS-algorithm |
| make_a_filter.c | Create filter for A-Weighting |
| read_obs.c | Data input/output routines for HMM programmes |
| rmsana.c | RMS histogram and statistics of sampled data and rms-data |
| rmsbed.c | RMS based begin-endpoint-detection |
| sad.c | Symmetric Adaptive Decorrelation |
| sel_list.c | Selection of data |
| ssp_master.c | Definitions for the signal processing routines |
| ssp_part1.c | Signal processing master routines |
| ssp_part2.c | Signal processing master routines |
| ssp_part3.c | Signal processing I/O routines and other dot functions |
| ssp_part4.c | Paste buffer for frames during signal processing |
| ssp_tbl.c | Link to all signal processing routines |
| sspdef.c | Initialisation of global variables |
| ssvdc.c | Svd (dummy) |
| window.c | Window part of a signal and create windows |
| sort | |
| qsortx.c | Sort an array while keeping several other array's consistent |
| sort.c | Sorting an array of elements using the quicksort algorithm |
| type | |
| endian.c | Detect byte order, and swap bytes in 2, 4 and 8 byte data structures |
| util | |
| bss_util.c | Utility routines |
| cmd.c | Simple command interpreter |
| indexdec.c | Index decoding |
| def.c | Provide the fabric (keywords, macro's, ...) that holds together the SPRAAK development environment |
| patch.c | Patches various routines and macro's to get them working on a specific platform |
| prog | |
| spr_add_gauss.c | Increase the level of tying in a reduced SC_HMM |
| spr_addkey.c | Add/modify a key in the header of a KEY file |
| spr_bmix.c | Mix the data in a file |
| spr_calc_lda.c | Calculate the lda transformation for a given preprocessing and segmentation file (classes) |
| spr_calc_oprobs.c | Dump state output probs |
| spr_copy.c | File copying with conversion |
| spr_ct2hmm.c | Combine the statistics from multiple segpass or vitpass runs into an updated model |
| spr_cwr_main.c | Large Vocabulary Continuous Speech Recognition with Hidden Markov models |
| spr_decorr.c | Calculate a good decorrelation matrix given the full covariance gaussians for an HMM-system with diagonal covariance matrices |
| spr_delkey.c | Delete a key in the header of a KEY file |
| spr_fv_gauss.c | Feature space MLLR |
| spr_fwdbwd_probs.c | Fwd-bwd alignment (viterbi alike) te determine boundary probabilities |
| spr_getkey.c | Get a key from the header of a KEY file |
| spr_hmmc2d.c | Convert a (semi)continuous density HMM to a discrete HMM or to a not-reduced semi-continuous HMM |
| spr_hmmcheck.c | Check the validity of an HMM |
| spr_hmmd2c.c | Convert a discrete HMM to a continuous or a semi-continuous HMM |
| spr_hmmmerge.c | Concatenate two HMMs |
| spr_hmmnormalize.c | Normalizes input file so that probabilities sum up to one |
| spr_hmmprint.c | Print the values in an HMM file |
| spr_hmmshake.c | Reorder the states in an HMM |
| spr_hmmsmooth.c | Smooth the mixture weights in HMM's retrained on little data |
| spr_hmmsum.c | Combine two HMMs by summing the moments |
| spr_htk2key.c | Convert from HTK format to the KEY format |
| spr_kappend.c | Append KEY files |
| spr_kdiff.c | Compare KEY files |
| spr_knn_dtw.c | Find the k nearest neighbour templates using DTW for a given context dependent phone graph |
| spr_lex_cvt.c | Modify a lexicon on a line-by-line basis |
| spr_lm_arpabo.c | Convert arpa style N-gram's to and from the binary format used in SPRAAK |
| spr_mdt_make_cgt.c | Create cluster gaussian to back-end gaussian association table |
| spr_mlp_bp.c | Training of MLP's |
| spr_msg_exec.c | Execute + time a program |
| spr_mvg_init.c | Initialization for K-means cluster centra and multi variate gaussians |
| spr_nn.c | |
| spr_printf.c | Test program |
| spr_scoreres.c | (Re)calculate the CSR score for a given result file |
| spr_segpass.c | Initialize an hmm file on the basis of unit or state level segmentations |
| spr_sel_frames.c | Select frames corresponding to units or states, or to all data |
| spr_sel_gauss.c | Reduce the number of Gaussians in a mixture |
| spr_setminprob.c | Set minimal probabilities in an HMM |
| spr_sigp.c | Signal processing |
| spr_space_cake.c | Continue Automatische Klank-Evaluatie |
| spr_spraak_lex.c | Speech to text converter using the SPRaak-interface |
| spr_spraak_socket.c | Speech interface using sockets |
| spr_spraak_STT.c | Speech to text converter using the SPRaak-interface |
| spr_tree2model.c | Construction of acoustic models given a decision tree |
| spr_tree_create.c | Creation of phonetic decision trees for context clustering |
| spr_tree_gauss_split.c | Decision tree based splitting of gaussians for HMMs with tied gaussians |
| spr_tt.c | Test program |
| spr_vitalign.c | Viterbi alignment |
| spr_vitpass.c | One pass of viterbi training for HMM-models |
| spr_vitprobs.c | Viterbi alignment + path probabilities |
| spr_wlatp.c | Processing of word lattices |
| scripts | |
| alien | |
| cmudict2yapa.py | |
| kaldi_cmvn_arc_to_corpus.py | |
| kaldi_feat_arc_to_spraak.py | |
| KaldiDNN2SPRAAK.py | |
| nbest_base.py | |
| spr_add_state_numbers.py | |
| spr_fsm2spr_fsg.py | |
| spr_pylib | |
| capi | |
| __init__.py | |
| bind_high_lvl_api.py | |
| clex.py | |
| core.py | |
| cstruct.py | |
| lex.py | |
| util.py | |
| io | |
| __init__.py | Tags spraak.io as a Python package |
| framefile.py | Provides code to read and write generic key and frame files |
| train | |
| __init__.py | Tags spraak.train as a Python package |
| clip_wgt.py | |
| cmd_names.py | Handles the conversion from the old (hmm75) executable to the new (SPRAAK) executables |
| concat_seg_files.py | |
| default_param_gi.py | |
| default_param_op.py | |
| extract_ci.py | |
| make_phtied_hmm.py | |
| make_tied_hmm.py | |
| make_tied_mvg.py | |
| make_triv_cd.py | |
| make_untied_hmm.py | |
| make_untied_mvg.py | |
| master.py | Defines the trainer and config objects that are available in the configuration file |
| ngauss_from_seg.py | |
| recovery.py | Handles recovering of old versions of resources |
| resource.py | Defines a generic resource and a specific implementation: a file resource |
| schedule.py | Defines schedulers for scheduling tasks |
| seg_ts2s.py | |
| task.py | Defines a generic task and all specific tasks derived from it |
| term.py | Provides support for colored terminal output, if available on the platform |
| util.py | Provides various utility functionalities, including logging and launching of external processes (local or remote) |
| wlat | |
| __init__.py | Tags spraak.wlat as a Python package |
| align2cor.py | |
| arc_sort.py | |
| best_end.py | |
| best_only.py | |
| check.py | |
| dens.py | |
| io.py | |
| lin_print.py | |
| renum_max.py | |
| rm_deps.py | |
| rm_deps2.py | |
| symb_xlat.py | |
| __init__.py | Tags spraak as a Python package |
| spr_eval.py | Script to perform evaluation experiments |
| spr_hmm_merge.py | ############################################################ __ _ _ _ _ # (_ |_) |_) |_| |_| |_/ Speech Processing, Recognition & # __) | | \ | | | | | \ Automatic Annotation Kit # # Copyright 2014 UGent # # Use of this software is governed by a License Agreement |
| spr_info.py | Extracts basic information from an hmm file |
| spr_make_pybind.py | Generates Python bindings for SPRAAK (advanced) |
| spr_mdt_make_cg.py | |
| spr_mdt_make_cgt.py | |
| spr_mdt_make_prospect_mat.py | |
| spr_mdt_make_vq.py | |
| spr_parser.py | Code parser to generate include header files and prepare documentation extraction |
| spr_pydoc_filter.py | Doxygen filter to handle custom comment formatting |
| spr_pydoc_indexer.py | |
| spr_train.py | Script to create and train acoustic models |
| spr_util.py | |
| spr_version.py | Set, modify or report the current version of SPRAAK |
| spr_wlat.py | Launch script for wlat processing scripts (experimental code) |
| spr_wlat2dot.py | Convert word lattice to dot graph file format |
| spr_wlat2sclite.py | Convert word lattice to nist sclite format |
| index.c | |
1.8.6
