構造体 wchmm_info
#include <wchmm.h>
wchmm_infoのコラボレーション図
変数 | |
| HTK_HMM_INFO * | hmminfo |
| HMM definitions used to construct this lexicon. | |
| NGRAM_INFO * | ngram |
| N-gram used to construct this lexicon. | |
| WORD_INFO * | winfo |
| Word dictionary used to construct this lexicon. | |
| int | maxwcn |
| Memory assigned maximum number of nodes. | |
| int | n |
| Num of nodes in this lexicon. | |
| WCHMM_STATE * | state |
| HMM state on tree lexicon [nodeID]. | |
| WORD_ID * | ststart |
| Word ID that begins at the state [nodeID]. | |
| WORD_ID * | stend |
| Word ID that ends at the state [nodeID]. | |
| int ** | offset |
| Node ID of a phone [wordID][0..phonelen-1]. | |
| int * | wordend |
| Node ID of word-end state [wordID]. | |
| int | startnum |
| Number of root nodes. | |
| int * | startnode |
| Root node index [0..startnum-1] -> node ID. | |
| int * | start2isolate |
| Root node index -> isolated root node ID. | |
| int | isolatenum |
| Number of isolated root nodes. | |
| LOGPROB * | wordend_a |
| Transition prob. outside word [wordID]. | |
| unsigned char * | outstyle |
| ID to indicate type of output probability container (one of AS_Style). | |
| S_CELL ** | sclist |
| List of successor list [scid]. | |
| int * | sclist2node |
| Mapping successor list [scid] to node. | |
| int | scnum |
| Number of factoring nodes that has successor list. | |
| LOGPROB * | fscore |
| List of 1-gram factoring score [-scid]. | |
| int | fsnum |
| Number of fscore. | |
| BMALLOC_BASE * | malloc_root |
| Pointer for block memory allocation. | |
説明
Whole lexicon tree structure holding all information.
この構造体の説明は次のファイルから生成されました:
- julius/wchmm.h
