adapt_am_t | A wrapper for all adaptation operations in Sphinx 3.X including MLLR and MAP |
align_phseg_s | |
align_stseg_s | |
align_wdseg_s | |
arraysize_t | Wrapper of array size |
ascr_t | |
backoff_t | |
backpointer_s | |
beam_t | Structure that contains all beam parameters for beam pruning in Viterbi algorithm |
bg32_t | A bigram structure which has 32 bits |
bg_t | A bigram structure |
blkarray_list_s | |
ca_dag_type | |
ca_daglink_type | |
ca_dagnode_type | |
ciphone_t | CI phone information |
class_t | Class to store the classifier parameters |
classifier_t | |
condition_t | |
conf_srch_hyp | |
conf_srch_hyp_t | Hypothesis structure that stores the confidence scores. Mainly used in confidence.c |
corpus_t | Structure for a corpus: essentially a set of strings each associated with a unique ID. Structure for a corpus: essentially a set of strings each associated with a unique ID. (Such as a reference sentence file, hypothesis file, and various control files.) NOTE: IDs are CASE-SENSITIVE |
cst_lts_rule_struct | |
cst_lts_rules_struct | |
ctxt_table_t | |
dag_t | |
daglink_s | |
daglink_t | |
dagnode_s | |
dict2pid_t | Building composite triphone (as well as word internal triphones) with the dictionary |
dict_t | Structure for a dictionary |
dictword_t | Structure for one dictionary word |
downsampling_t | Structure containing various parameters for downsampling |
endpointer_t | Struct to hold the end-pointer parameters |
endptr_t | |
fast_gmm_t | Structure that contains all parameter related to 4-Level Fast GMM computation |
filler2phnMap | |
fillpen_t | |
fpoint_t | |
fsg_hist_entry_s | |
fsg_history_s | |
fsg_lextree_s | |
fsg_pnode_ctxt_t | |
fsg_pnode_s | Fsg node. All transitions (words) out of any given FSG state represented are by a phonetic prefix lextree (except for epsilon or null transitions; they are not part of the lextree). Lextree leaf nodes represent individual FSG transitions, so no sharing is allowed at the leaf nodes. The FSG transition probs are distributed along the lextree: the prob at a node is the max of the probs of all leaf nodes (and, hence, FSG transitions) reachable from that node |
fsg_search_s | |
fwd_dbg_t | |
gau_select_t | Structure that contains all parameters related to Gaussian selection |
gauden_dist_t | Structure to store distance (density) values for a given input observation wrt density values in some given codebook |
gauden_t | Multivariate gaussian mixture density parameters |
gmm_select_t | Structure that contains all parameters for CI-based GMM selection |
grp_str_t | |
gs_s | |
gs_t | |
history_s | |
histprune_t | Structure containing various histogram pruning parameters and internal storage All in integers |
hmm_context_s | |
hmm_context_t | Shared information between a set of HMMs |
hmm_s | |
hmm_state_t | A single state in the HMM |
hmm_t | An individual HMM among the HMM search space |
hyp_t | Alias of srch_hyp_t; |
interp_t | Wrapper structure of CD/CI interpolation |
interp_t::interp_wt_s | Quick and Dirty wrapper of the CD and CI weight |
kb_t | |
kbcore_t | |
kd_tree_node_s | |
kd_tree_s | |
lattice_s | |
lattice_t | |
latticehist_t | |
lex_entry_str | |
lextree_lcroot_t | |
lextree_node_t | |
lextree_t | |
lm_s | |
lm_t | The language model. All unigrams are read into memory on initialization. Bigrams and trigrams read in on demand |
lm_tgcache_entry32_t | |
lm_tgcache_entry_t | |
lmclass_s | |
lmclass_set_s | |
lmclass_set_t | Collection of LM classes. Most applications would use multiple classes. This data type is provided as a convenience for maintaining several such classes |
lmclass_t | An LM class object |
lmclass_word_s | |
lmclass_word_t | A single word in an LM class |
lmlog_t | Log quantities represented in either floating or integer format |
lmset_s | |
lmset_t | Structure for multiple LM, provide operations for addition/deletion/read Structure for multiple, named LMs, started from s2 |
mdef_t | Strcture for storing the model definition |
membg32_t | A 32 bits version of membg_t |
membg_t | Management of in-memory bigrams. Not used if all bigrams in memory |
mgau2sen_s | |
mgau_model_t | The set of mixture-Gaussians in an acoustic model used in Sphinx 3.X family of tool |
mgau_t | A single mixture-Gaussian model for one senone (see above comment) |
ms_mgau_model_t | |
ms_mgau_t | Multi-stream mixture gaussian. It is not necessary to be continr |
node_s | |
ph_lc_s | |
ph_lc_t | Structures for storing the left context |
ph_rc_s | |
ph_rc_t | Structures needed for mapping <ci,lc,rc,wpos> into pid. (See mdef_t.wpos_ci_lclist below.) (lc = left context; rc = right context.) NOTE: Both ph_rc_t and ph_lc_t FOR INTERNAL USE ONLY |
phone_t | Triphone information, including base phones as a subset. For the latter, lc, rc and wpos are non-existent |
pl_t | Structure that contains all parameters for phoneme lookahead |
plink_s | |
pnode_s | |
point_t | Wrapper of a point in integer |
s2_fsg_s | |
s2_fsg_trans_s | |
s2_semi_mgau_s | |
s3_arraylist_s | |
s3_cfg_entry_s | |
s3_cfg_item_t | |
s3_cfg_rule_s | |
s3_cfg_state_s | |
s3_cfg_t | |
s3_decode_t | |
s3_endpointer_t | |
seg_hyp_line | |
seg_hyp_line_t | Strurcture that stores one line of hypothesis. Mainly used in confidence.c |
senone_t | 8-bit senone PDF structure |
slink_s | |
snode_s | |
sort_array_t | |
sort_t | |
sorted_entry_s | |
sorted_entry_t | Single entry used in the linked list structure of lm reading |
sorted_list_t | The sorted list used lm reading. list is a (64K long) array. The first entry is the root of the tree and is created during initialization |
srch_FLAT_FWD_graph_s | |
srch_funcs_s | |
srch_hyp_s | |
srch_hyp_t | Hypothesis structure |
srch_s | |
srch_WST_graph_t | |
stat_t | Structure to hold all statistics in Sphinx 3.x |
subvq_t | Wrapper structures of sub-vector quantization |
tg32_t | A 32 bits version of tg_t |
tg_t | A trigram structure |
tginfo32_s | |
tginfo32_t | 32 bit version of tginfo |
tginfo_s | |
tginfo_t | Trigram cache that enhance locating trigram for a given bigram (w_1,w_2) |
tmat_t | Transition matrix data structure. All phone HMMs are assumed to have the same topology |
ug_t | A unigram structure Please see |
utt_res_t | A structure to store utterance-based resource Assume that most resource are string pointers, the string itself is pre-allocated somewhere |
vector_gautbl_t | |
vh_lms2vh_t | |
vh_lmstate_u | |
vithist_entry_t | |
vithist_t | |
vqFeature_t | |
whmm_s | |
whmm_t | Word HMM instance: the basic structure searched during recognition |
word_cand_s | |
word_fsg_s | |
word_fsglink_s | |
word_graph_link_t | |
word_graph_node_t | |
word_graph_t | |
word_ugprob_s | |
word_ugprob_t | |
wordprob_t | Generic structure that could be used at any n-gram level |
xwdssid_t | Cross word triphone model structure |