#include <search.h>
Collaboration diagram for __node__:
Data Fields | |
| __node__ * | next |
| Link to next hypothesis, used in stack. | |
| __node__ * | prev |
| Link to previous hypothesis, used in stack. | |
| boolean | endflag |
| TRUE if this is a final sentence result. | |
| WORD_ID | seq [MAXSEQNUM] |
| Word sequence. | |
| short | seqnum |
| Length of seq. | |
| LOGPROB | score |
| Total score (forward+backward, LM+AM). | |
| short | bestt |
| Best connection frame of last word in word trellis. | |
| short | estimated_next_t |
| Estimated next connection time frame (= beginning of last word on word trellis): next word hypothesis will be looked up near this frame on word trellis. | |
| LOGPROB * | g |
| Current forward viterbi score in each frame. | |
| LOGPROB | final_g |
| Extra forward score on end of frame for multipath mode. | |
| int | state |
| (dfa) Current DFA state ID | |
| TRELLIS_ATOM * | tre |
| Trellis word of last word. | |
| LOGPROB * | g_prev |
| Viterbi score back to last 1 phoneme. | |
| HMM_Logical * | last_ph |
| Last applied triphone. | |
| boolean | last_ph_sp_attached |
| Last phone which the inter-word sp has been attached for multipath mode. | |
| LOGPROB | lscore |
| N-gram score of last word (will be used for 1-phoneme backscan and graph output, always 0 for dfa. | |
| LOGPROB | totallscore |
| (n-gram) Accumulated language score (LM only) | |
| LOGPROB | cmscore [MAXSEQNUM] |
| Confidence score of each word. | |
| WordGraph * | prevgraph |
| Graph word corresponding to the last word. | |
| WordGraph * | lastcontext |
| Graph word of next previous word. | |
| LOGPROB | tail_g_score |
| forward g score for later score adjustment | |
| __recogprocess__ * | region |
Definition at line 88 of file search.h.
1.5.1