#include <recog.h>
Collaboration diagram for __StackDecode__:
Data Fields | |
int | hypo_len_count [MAXSEQNUM+1] |
Count of popped hypothesis per each length. | |
int | maximum_filled_length |
Current least beam-filled depth. | |
LOGPROB * | framemaxscore |
Maximum score of each frame on 2nd pass for score enveloping. | |
NODE * | stocker_root |
Node stocker for recycle. | |
int | popctr |
Num of popped hypotheses from stack. | |
int | genectr |
Num of generated hypotheses. | |
int | pushctr |
Num of hypotheses actually pushed to stack. | |
int | finishnum |
Num of found sentence hypothesis. | |
NODE * | current |
Current node for debug. | |
LOGPROB | cm_alpha |
alpha scaling value from jconf | |
LOGPROB * | wordtrellis [2] |
Buffer to compute viterbi path of a word. | |
LOGPROB * | g |
Buffer to hold source viterbi scores. | |
HMM_Logical ** | phmmseq |
Phoneme sequence to be computed. | |
int | phmmlen_max |
Maximum length of phmmseq. | |
boolean * | has_sp |
Mark which phoneme allow short pause for multi-path mode. |
Definition at line 194 of file recog.h.