pass1.c

julius version 4.12.about sound recognition.

/**
 * @file   pass1.c
 * 
 * <JA>
 * @brief  妈1パス¨フレ〖ム票袋ビ〖ム玫瑚
 *
 * 琅弄腾菇陇辑今を脱いて·掐蜗泼魔翁ベクトル误に滦して·Juliusの妈１パス
 * であるフレ〖ム票袋ビ〖ム玫瑚を乖います. 
 *
 * 掐蜗デ〖タ链挛があらかじめ评られている眷圭は·办崇で纷换を
 * 乖う簇眶 get_back_trellis() がメインから钙ばれます. オンライン千急
 * の眷圭は realtime_1stpass.c から·介袋步·フレ〖ムごとの纷换·
 * 姜位借妄のそれぞれが掐蜗の渴乖にあわせて改侍に钙ばれます. 
 *
 * 悸狠の改」の千急借妄インスタンスごとの借妄は beam.c に淡揭されています. 
 *
 * </JA>
 * 
 * <EN>
 * @brief  The first pass: frame-synchronous beam search
 *
 * These functions perform a frame-synchronous beam search using a static
 * lexicon tree, as the first pass of Julius/Julian.
 *
 * When the whole input is already obtained, get_back_trellis() simply
 * does all the processing of the 1st pass.  When performing online
 * real-time recognition with concurrent speech input, each function
 * will be called separately from realtime_1stpass.c according on the
 * basis of input processing.
 *
 * The core recognition processing functions for each recognition
 * process instances are written in beam.c.
 *
 * </EN>
 * 
 * @author Akinobu Lee
 * @date   Fri Oct 12 23:14:13 2007
 *
 * $Revision: 1.7 $
 * 
 */
/*
 * Copyright (c) 1991-2007 Kawahara Lab., Kyoto University
 * Copyright (c) 2000-2005 Shikano Lab., Nara Institute of Science and Technology
 * Copyright (c) 2005-2007 Julius project team, Nagoya Institute of Technology
 * All rights reserved
 */

#include <julius/julius.h>

/********************************************************************/
/* 妈１パスを悸乖するメイン簇眶                                     */
/* 掐蜗をパイプライン借妄する眷圭は realtime_1stpass.c を徊救のこと */
/* main function to execute 1st pass                                */
/* the pipeline processing is not here: see realtime_1stpass.c      */
/********************************************************************/

/** 
 * <EN>
 * @brief  Process one input frame for all recognition process instance.
 *
 * This function proceeds the recognition for one frame.  All
 * recognition process instance will be processed synchronously.
 * The input frame for each instance is stored in mfcc->f, where mfcc
 * is the MFCC calculation instance assigned to each process instance.
 *
 * If an instance's mfcc->invalid is set to TRUE, its processing will
 * be skipped.
 *
 * When using GMM, GMM computation will also be executed here.
 * If GMM_VAD is defined, GMM-based voice detection will be performed
 * inside this function, by using a scheme of short-pause segmentation.
 *
 * This function also handles segmentation of recognition process.  A
 * segmentation will occur when end of speech is detected by level-based
 * sound detection or GMM-based / decoder-based VAD, or by request from
 * application.  When segmented, it stores current frame and return with
 * that status.
 *
 * The frame-wise callbacks will be executed inside this function,
 * when at least one valid recognition process instances exists.
 * 
 * </EN>
 * <JA>
 * @brief  链ての千急借妄インスタンス借妄を1フレ〖ム尸渴める.
 *
 * 链ての千急借妄インスタンスについて·充り烧けられているMFCC纷换インスタンス
 * の mfcc->f をカレントフレ〖ムとして借妄を1フレ〖ム渴める. 
 *
 * なお·mfcc->invalid が TRUE となっている借妄インスタンスの借妄はスキップ
 * される. 
 *
 * GMMの纷换もここで钙び叫される. GMM_VAD 年盗箕は·GMM による
 * 券厦惰粗倡幌ˇ姜位の浮叫がここで乖われる. また·GMMの纷换冯蔡·
 * あるいは千急借妄柒のショ〖トポ〖ズセグメンテ〖ション冉年やデバイスˇ嘲婶
 * からの妥滇によりセグメンテ〖ションが妥滇されたかどうかの冉年も乖う. 
 *
 * フレ〖ム帽疤で钙び叫されるコ〖ルバックが判峡されている眷圭は·それらの
 * 钙叫しも乖う. 
 * </JA>
 * 
 * @param recog [in] engine instance
 * 
 * @return 0 on success, -1 on error, or 1 when an input segmentation
 * occured/requested inside this function.
 *
 * @callgraph
 * @callergraph
 * 
 */
int
decode_proceed(Recog *recog)
{
  MFCCCalc *mfcc;
  boolean break_flag;
  boolean break_decode;
  RecogProcess *p;
  boolean ok_p;
#ifdef GMM_VAD
  GMMCalc *gmm;
  boolean break_gmm;
#endif
  
  break_decode = FALSE;

  for(p = recog->process_list; p; p = p->next) {
#ifdef DETERMINE
    p->have_determine = FALSE;
#endif
    p->have_interim = FALSE;
  }
  for (mfcc = recog->mfcclist; mfcc; mfcc = mfcc->next) {
    mfcc->segmented = FALSE;
  }

#ifdef POWER_REJECT
  for (mfcc = recog->mfcclist; mfcc; mfcc = mfcc->next) {
    if (!mfcc->valid) continue;
    if (mfcc->f == 0) {
      mfcc->avg_power = 0.0;
      if (debug2_flag) jlog("STAT: power_reject: reset\n");
    }
  }
#endif


#ifdef GMM_VAD
  if (recog->gmm != NULL) {
    /* reset flags */
    break_gmm = FALSE;
    recog->gc->want_rewind = FALSE;
  }
#endif
  if (recog->gmm != NULL && recog->gmmmfcc->valid) {
    /* GMM 纷换を乖う */
    if (recog->gmmmfcc->f == 0) {
      /* GMM 纷换の介袋步 */
      gmm_prepare(recog);
    }
    /* このフレ〖ムに滦するGMMの锑刨を纷换 */
    gmm_proceed(recog);
#ifdef GMM_VAD
    /* Check for GMM-based VAD */
    gmm = recog->gc;
    gmm_check_trigger(recog);
    if (gmm->after_trigger) {
      /* after trigger, in speech area */
      if (gmm->down_trigger) {
	/* down trigger, end segment */
#ifdef GMM_VAD_DEBUG
	printf("GMM_VAD: %d: down trigger\n", recog->gmmmfcc->f);
#endif
	recog->gmmmfcc->sparea_start = recog->gmmmfcc->f + 1 - recog->jconf->detect.gmm_margin;
	if (recog->gmmmfcc->sparea_start < 0) recog->gmmmfcc->sparea_start = 0;
	gmm->after_trigger = FALSE;
	recog->gmmmfcc->segmented = TRUE;
	break_gmm = TRUE;
      } else {
	/* keep recognition */
      }
    } else {
      /* before trigger, in noise area */
      if (gmm->up_trigger) {
	/* start recognition */
	/* request caller to rewind to the backstep point and
	   re-start with normal search */
	if (recog->gmmmfcc->f + 1 < recog->jconf->detect.gmm_margin) {
	  gmm->rewind_frame = 0;
	} else {
	  gmm->rewind_frame = recog->gmmmfcc->f + 1 - recog->jconf->detect.gmm_margin;
	}
#ifdef GMM_VAD_DEBUG
	printf("GMM_VAD: %d: up trigger, start recognition with %d frame rewind\n", recog->gmmmfcc->f, recog->gmmmfcc->f - gmm->rewind_frame);
#endif
	gmm->want_rewind = TRUE;
	gmm->want_rewind_reprocess = TRUE;
	gmm->after_trigger = TRUE;
	return 0;
      } else {
	/* before trigger, noise continues */

	/* if noise goes more than a certain frame, shrink the noise area
	   to avoid unlimited memory usage */
	if (recog->gmmmfcc->f + 1 > GMM_VAD_AUTOSHRINK_LIMIT) {
	  gmm->want_rewind = TRUE;
	  gmm->want_rewind_reprocess = FALSE;
	  gmm->rewind_frame = recog->gmmmfcc->f + 1 - recog->jconf->detect.gmm_margin;
	  if (debug2_flag) {
	    jlog("DEBUG: GMM_VAD: pause exceeded %d, rewind\n", GMM_VAD_AUTOSHRINK_LIMIT);
	  }
	}

	/* skip recognition processing */
	return 0;
      }
    }
#endif /* GMM_VAD */
  }

  for(p = recog->process_list; p; p = p->next) {
    if (!p->live) continue;
    mfcc = p->am->mfcc;
    if (!mfcc->valid) {
      /* このフレ〖ムの借妄をスキップ */
      /* skip processing the frame */
      continue;
    }

    /* mfcc-f のフレ〖ムについて千急借妄(フレ〖ム票袋ビ〖ム玫瑚)を渴める */
    /* proceed beam search for mfcc->f */
    if (mfcc->f == 0) {
      /* 呵介のフレ〖ム: 玫瑚借妄を介袋步 */
      /* initial frame: initialize search process */
      if (get_back_trellis_init(mfcc->param, p) == FALSE) {
	jlog("ERROR: %02d %s: failed to initialize the 1st pass\n", p->config->id, p->config->name);
	return -1;
      }
    }
    if (mfcc->f > 0 || p->am->hmminfo->multipath) {
      /* 1フレ〖ム玫瑚を渴める */
      /* proceed search for 1 frame */
      if (get_back_trellis_proceed(mfcc->f, mfcc->param, p, FALSE) == FALSE) {
	mfcc->segmented = TRUE;
	break_decode = TRUE;
      }
      if (p->config->successive.enabled) {
	if (detect_end_of_segment(p, mfcc->f - 1)) {
	  /* セグメント姜位浮梦: 妈１パスここで面们 */
	  mfcc->segmented = TRUE;
	  break_decode = TRUE;
	}
      }
    }
  }

  /* セグメントすべきかどうか呵姜弄な冉年を乖うˉ
     デコ〖ダベ〖スVADあるいは spsegment の眷圭·剩眶インスタンス粗で OR
     を艰るˉまた·GMMなど剩眶答洁がある眷圭は答洁粗で AND を艰るˉ*/
  /* determine whether to segment at here
     If multiple segmenter exists, take their AND */
  break_flag = FALSE;
  if (break_decode
#ifdef GMM_VAD
      || (recog->gmm != NULL && break_gmm)
#endif
      ) {
    break_flag = TRUE;
  }

  if (break_flag) {
    /* 玫瑚借妄の姜位が券栏したのでここで千急を姜える. 
       呵介のフレ〖ムから [f-1] 戎誊までが千急されたことになる
    */
    /* the recognition process tells us to stop recognition, so
       recognition should be terminated here.
       the recognized data are [0..f-1] */

    /* 呵姜フレ〖ムを last_time にセット */
    /* set the last frame to last_time */
    for (mfcc = recog->mfcclist; mfcc; mfcc = mfcc->next) {
      mfcc->last_time = mfcc->f - 1;
    }

    if (! recog->jconf->decodeopt.segment) {
      /* ショ〖トポ〖ズ笆嘲で磊れた眷圭·荒りのサンプルは千急せずに嘉てる */
      /* drop rest inputs if segmented by error */
      for (mfcc = recog->mfcclist; mfcc; mfcc = mfcc->next) {
	mfcc->param->header.samplenum = mfcc->f;
	mfcc->param->samplenum = mfcc->f;
      }
    }

    return 1;
  }

  /* call frame-wise callback for the processing results if any */
#ifdef DETERMINE
  ok_p = FALSE;
  for(p=recog->process_list;p;p=p->next) {
    if (!p->live) continue;
    if (p->have_determine) {
      ok_p = TRUE;
    }
  }
  if (ok_p) callback_exec(CALLBACK_RESULT_PASS1_DETERMINED, recog);
#endif
  ok_p = FALSE;
  for(p=recog->process_list;p;p=p->next) {
    if (!p->live) continue;
    if (p->have_interim) {
      ok_p = TRUE;
    }
  }
  if (ok_p) callback_exec(CALLBACK_RESULT_PASS1_INTERIM, recog);
  
  return 0;
}

#ifdef POWER_REJECT
boolean
power_reject(Recog *recog)
{
  MFCCCalc *mfcc;

  for (mfcc = recog->mfcclist; mfcc; mfcc = mfcc->next) {
    /* skip if not realtime and raw file processing */
    if (mfcc->avg_power == 0.0) continue;
    if (debug2_flag) jlog("STAT: power_reject: MFCC%02d: avg_power = %f\n", mfcc->id, mfcc->avg_power / mfcc->param->samplenum);
    if (mfcc->avg_power / mfcc->param->samplenum < recog->jconf->reject.powerthres) return TRUE;
  }
  return FALSE;
}
#endif

/** 
 * <EN>
 * @brief  End procedure of the first pass (when segmented)
 *
 * This function do things for ending the first pass and prepare for
 * the next recognition, when the input was segmented at the middle of
 * recognition by some reason.
 *
 * First, the best path at each recognition process instance will be parsed
 * and stored.  In case of recognition error or input rejection, the error
 * status will be set.
 *
 * Then, the last pause segment of the processed input will be cut and saved
 * to be processed at first in the recognition of the next or remaining input.
 * 
 * </EN>
 * <JA>
 * @brief  妈1パスの姜位借妄∈セグメント箕∷
 * 
 * 掐蜗が部らかの祸统によって庞面でセグメントされた箕に·妈1パスの千急借妄を
 * 姜位して肌搀浩倡するための借妄を乖う. 
 *
 * まず·称千急借妄インスタンスに滦して·呵锑帽胳废误を斧烧け·妈1パスの
 * 千急冯蔡として呈羌する. また·千急己窃ˇ掐蜗逮笛の箕はエラ〖ステ〖タスをそ
 * れぞれセットする.
 * 
 * そして·肌搀の千急で·肌のセグメントの千急を·浮叫された琐萨花不
 * 惰粗から浩倡するために·その琐萨花不惰粗を磊り叫しておく借妄を钙ぶ. 
 *