int recordInit()
{
ad = ad_open();
if (ad == NULL) {
printf("Error opening recording device.\n");
return 0;
}
c_ad = cont_ad_init(ad, ad_read);
if (c_ad == NULL) {
printf("Error initializing continues ad.\n");
return 0;
}
if (ad_start_rec(ad) < 0) {
printf("Error starting recording.\n");
return 0;
}
if (cont_ad_calib(c_ad) < 0) {
printf("Error calibrating continues ad.\n");
return 0;
}
ad_stop_rec(ad);
return 1;
}
int
main(int argc, char *argv[])
{
cont_ad_t *cont;
ad_rec_t ad;
int16 buf[512];
int listening;
ad.sps = 16000;
ad.bps = 2;
TEST_ASSERT(infp = fopen(TESTDATADIR "/chan3.raw", "rb"));
TEST_ASSERT(cont = cont_ad_init(&ad, file_ad_read));
printf("Calibrating ...");
fflush(stdout);
if (cont_ad_calib(cont) < 0)
printf(" failed; file too short?\n");
else
printf(" done after %ld samples\n", ftell(infp) / 2);
rewind(infp);
listening = FALSE;
while (1) {
int k = cont_ad_read(cont, buf, 512);
/* End of file. */
if (k < 0) {
if (listening) {
printf("End of file at %.3f seconds\n",
(double)(cont->read_ts - k) / 16000);
}
break;
}
if (cont->state == CONT_AD_STATE_SIL) {
/* Has there been enough silence to cut the utterance? */
if (listening && cont->seglen > 8000) {
printf("End of utterance at %.3f seconds\n",
(double)(cont->read_ts - k - cont->seglen) / 16000);
listening = FALSE;
}
}
else {
if (!listening) {
printf("Start of utterance at %.3f seconds\n",
(double)(cont->read_ts - k) / 16000);
listening = TRUE;
}
}
}
cont_ad_close(cont);
fclose(infp);
return 0;
}
static gboolean
gst_sphinx_sink_start (GstBaseSink * asink)
{
GstSphinxSink *sphinxsink = GST_SINK (asink);
sphinxsink->ad.self = sphinxsink;
sphinxsink->ad.sps = 16000;
sphinxsink->ad.self = sphinxsink;
sphinxsink->ad.bps = sizeof(int16);
sphinxsink->ad.calibrated = FALSE;
sphinxsink->ad.calibrate_started = FALSE;
sphinxsink->cont = cont_ad_init ((ad_rec_t*)&sphinxsink->ad, gst_sphinx_sink_ad_read);
return TRUE;
}
static gboolean
gst_pocketsphinx_start (GstBaseSink * asink)
{
GstPocketSphinx *sphinxsink = GST_POCKETSPHINX (asink);
GstSphinxSinkAd *ad = &sphinxsink->ad;
ad->self = sphinxsink;
ad->sps = 16000;
ad->self = sphinxsink;
ad->bps = sizeof(int16);
ad->calibrated = FALSE;
ad->calibrate_started = FALSE;
sphinxsink->cont = cont_ad_init ((ad_rec_t *)ad, gst_pocketsphinx_ad_read);
return TRUE;
}
int init()
{
cmd_ln_t *config;
printf("Initializing audio device...\n");
config = cmd_ln_init(NULL, ps_args(), TRUE,
"-hmm", MODELDIR "/hmm/en_US/hub4wsj_sc_8k",
"-lm", MODELDIR "/lm/en/turtle.DMP",
"-dict", MODELDIR "/lm/en/turtle.dic",
NULL);
if(config == NULL) {
fprintf(stderr, "Could not init sphinx config.\n");
return 1;
}
psDecoder = ps_init(config);
if(psDecoder == NULL) {
fprintf(stderr, "Could not init psDecoder.\n");
return 2;
}
audioDevice = ad_open();
if(audioDevice == NULL) {
fprintf(stderr, "Could not open Audio Device.\n");
return 3;
}
continousAudoDevice = cont_ad_init(audioDevice, ad_read);
if(continousAudoDevice == NULL) {
fprintf(stderr, "Could not open Audio Device.\n");
return 4;
}
if(ad_start_rec(audioDevice) < 0) {
fprintf(stderr,"Failed to start recording.\n");
return 5;
}
if(cont_ad_calib(continousAudoDevice) < 0) {
fprintf(stderr,"Failed to calibrate voice activity detection.\n");
return 6;
}
return 0;
}
static int openAudioDevice(inputThread_t *p_thread)
{
int ret;
// open audio device, used for recording audio data
p_thread->audioDevice = ad_open();
if (p_thread->audioDevice == NULL) {
PRINT_ERR("Failed to open audio device.\n");
return -1;
}
// init audio device as continous audio device
p_thread->contAudioDevice = cont_ad_init(p_thread->audioDevice, ad_read);
if (p_thread->contAudioDevice == NULL) {
PRINT_ERR("Failed to init continuous audio device.\n");
return -2;
}
// calibrate audio device
ret = ad_start_rec(p_thread->audioDevice);
if (ret < 0) {
PRINT_ERR("Failed to start recording (%d).\n", ret);
return ret;
}
ret = cont_ad_calib(p_thread->contAudioDevice);
if (ret < 0) {
PRINT_ERR("Failed to calibrate continuous audio device (%d).\n", ret);
return ret;
}
ad_stop_rec(p_thread->audioDevice);
pthread_barrier_wait(&p_thread->startBarrier);
return 0;
}
/*
* Main utterance processing loop:
* for (;;) {
* wait for start of next utterance;
* decode utterance until silence of at least 1 sec observed;
* print utterance result;
* }
*/
static void
recognize_from_microphone()
{
ad_rec_t *ad;
int16 adbuf[4096];
int32 k, ts, rem;
char const *hyp;
char const *uttid;
cont_ad_t *cont;
char word[256];
char c1[256], c2[256];
int tracking = 0;
int halted = 0;
int LEFT = 0;
int RIGHT = 1;
int MOVE_CENT = 100; //1 meter
int numwords;
setlinebuf(stdout);
if ((ad = ad_open_dev(cmd_ln_str_r(config, "-adcdev"),
(int)cmd_ln_float32_r(config, "-samprate"))) == NULL)
E_FATAL("Failed to open audio device\n");
/* Initialize continuous listening module */
if ((cont = cont_ad_init(ad, ad_read)) == NULL)
E_FATAL("Failed to initialize voice activity detection\n");
if (ad_start_rec(ad) < 0)
E_FATAL("Failed to start recording\n");
if (cont_ad_calib(cont) < 0)
E_FATAL("Failed to calibrate voice activity detection\n");
printf("LEDON BLUE\n");
for (;;) {
/* Indicate listening for next utterance */
fprintf(stderr, "READY....\n");
fflush(stderr);
/* Wait data for next utterance */
while ((k = cont_ad_read(cont, adbuf, 4096)) == 0)
sleep_msec(100);
if (k < 0)
E_FATAL("Failed to read audio\n");
/*
* Non-zero amount of data received; start recognition of new utterance.
* NULL argument to uttproc_begin_utt => automatic generation of utterance-id.
*/
if (ps_start_utt(ps, NULL) < 0)
E_FATAL("Failed to start utterance\n");
ps_process_raw(ps, adbuf, k, FALSE, FALSE);
fprintf(stderr, "Listening...\n");
/* Note timestamp for this first block of data */
ts = cont->read_ts;
/* Decode utterance until end (marked by a "long" silence, >1sec) */
for (;;) {
/* Read non-silence audio data, if any, from continuous listening module */
if ((k = cont_ad_read(cont, adbuf, 4096)) < 0)
E_FATAL("Failed to read audio\n");
if (k == 0) {
/*
* No speech data available; check current timestamp with most recent
* speech to see if more than 1 sec elapsed. If so, end of utterance.
*/
if ((cont->read_ts - ts) > DEFAULT_SAMPLES_PER_SEC)
break;
}
else {
/* New speech data received; note current timestamp */
ts = cont->read_ts;
}
/*
* Decode whatever data was read above.
*/
rem = ps_process_raw(ps, adbuf, k, FALSE, FALSE);
/* If no work to be done, sleep a bit */
if ((rem == 0) && (k == 0))
sleep_msec(20);
}
/*
* Utterance ended; flush any accumulated, unprocessed A/D data and stop
* listening until current utterance completely decoded
*/
ad_stop_rec(ad);
while (ad_read(ad, adbuf, 4096) >= 0);
cont_ad_reset(cont);
fprintf(stderr, "Stopped listening, please wait...\n");
fflush(stdout);
/* Finish decoding, obtain and print result */
ps_end_utt(ps);
hyp = ps_get_hyp(ps, NULL, &uttid);
fprintf(stderr, "%s: %s\n", uttid, hyp);
/* Exit if the first word spoken was GOODBYE */
if (hyp) {
numwords = sscanf(hyp, "%s %s %s", word, c1, c2);
if(strcmp(word, "GUGGUG") == 0) {
if(strcmp(c1, "HALT") == 0) {
printf("LEDOFF BLUE\n");
halted = 1;
} else if(strcmp(c1, "RESUME") == 0) {
printf("LEDON BLUE\n");
halted = 0;
}
if(strcmp(c1, "BEGIN") == 0 || strcmp(c1, "START") == 0) {
if(strcmp(c2, "TRACKING") == 0 && !tracking) {
printf("START TRACKING\n");
tracking = 1;
halted = 0;
}
} else if(strcmp(c1, "STOP") == 0) {
if(strcmp(c2, "TRACKING") == 0 && tracking) {
printf("STOP TRACKING\n");
tracking = 0;
}
}
if(!tracking && !halted && numwords == 3) {
if(strcmp(c1, "TURN") == 0) {
if(strcmp(c2, "AROUND") == 0) {
printf("TURN %d 180\n", LEFT);
} else if(strcmp(c2, "LEFT") == 0) {
printf("TURN %d 90\n", LEFT);
} else if(strcmp(c2, "RIGHT") == 0) {
printf("TURN %d 90\n", RIGHT);
}
} else if(strcmp(c1, "MOVE") == 0) {
if(strcmp(c2, "FORWARD") == 0) {
printf("MOVE 0 %d\n", MOVE_CENT);
} else if(strcmp(c2, "BACKWARD") == 0) {
printf("MOVE 1 %d\n", MOVE_CENT);
}
}
}
}
}
/* Resume A/D recording for next utterance */
if (ad_start_rec(ad) < 0)
E_FATAL("Failed to start recording\n");
}
cont_ad_close(cont);
ad_close(ad);
}
/*
* Continuous recognition from a file
*/
static void
recognize_from_file() {
cont_ad_t *cont;
ad_rec_t file_ad = {0};
int16 adbuf[4096];
const char* hyp;
const char* uttid;
int32 k, ts, start;
char waveheader[44];
if ((rawfd = fopen(cmd_ln_str_r(config, "-infile"), "rb")) == NULL) {
E_FATAL_SYSTEM("Failed to open file '%s' for reading",
cmd_ln_str_r(config, "-infile"));
}
fread(waveheader, 1, 44, rawfd);
file_ad.sps = (int32)cmd_ln_float32_r(config, "-samprate");
file_ad.bps = sizeof(int16);
if ((cont = cont_ad_init(&file_ad, ad_file_read)) == NULL) {
E_FATAL("Failed to initialize voice activity detection");
}
if (cont_ad_calib(cont) < 0)
E_FATAL("Failed to calibrate voice activity detection\n");
rewind (rawfd);
for (;;) {
while ((k = cont_ad_read(cont, adbuf, 4096)) == 0);
if (k < 0) {
break;
}
if (ps_start_utt(ps, NULL) < 0)
E_FATAL("ps_start_utt() failed\n");
ps_process_raw(ps, adbuf, k, FALSE, FALSE);
ts = cont->read_ts;
start = ((ts - k) * 100.0) / file_ad.sps;
for (;;) {
if ((k = cont_ad_read(cont, adbuf, 4096)) < 0)
break;
if (k == 0) {
/*
* No speech data available; check current timestamp with most recent
* speech to see if more than 1 sec elapsed. If so, end of utterance.
*/
if ((cont->read_ts - ts) > DEFAULT_SAMPLES_PER_SEC)
break;
}
else {
/* New speech data received; note current timestamp */
ts = cont->read_ts;
}
ps_process_raw(ps, adbuf, k, FALSE, FALSE);
}
ps_end_utt(ps);
if (cmd_ln_boolean_r(config, "-time")) {
print_word_times(start);
} else {
hyp = ps_get_hyp(ps, NULL, &uttid);
fprintf(stderr, "%s: %s\n", uttid, hyp);
}
fflush(stdout);
}
cont_ad_close(cont);
fclose(rawfd);
}
void listen::recognize_from_microphone(){
ad_rec_t *ad;
int16 adbuf[4096];
int32 k, ts, rem;
char buffer[128];
char const *hyp;
char const *uttid;
cont_ad_t *cont;
state = SLEEPING;
FILE* pipe = popen(c.getValue("[General]", "Hcidump").c_str(), "r");
std::string bufferStr;
std::size_t found;
if((ad = ad_open_dev(cmd_ln_str_r(config, "-adcdev"), (int)cmd_ln_float32_r(config, "-samprate"))) == NULL)
E_FATAL("Failed to open audio device\n");
/* Initialize continuous listening module */
if((cont = cont_ad_init(ad, ad_read)) == NULL)
E_FATAL("Failed to initialize voice activity detection\n");
if(ad_start_rec(ad) < 0)
E_FATAL("Failed to start recording\n");
if(cont_ad_calib(cont) < 0)
E_FATAL("Failed to calibrate voice activity detection\n");
while(!feof(pipe) || state == SLEEPING){
fgets(buffer, 128, pipe);
bufferStr = buffer;
found = bufferStr.find(c.getValue("[General]", "KeyPress"));
if(found!=std::string::npos){
i.pauseIfPlaying();
s.speakThis(c.getValue("[General]", "WakeUpPhrase"));
state = ACTIVE;
while(state != SLEEPING){
/* Indicate listening for next utterance */
printf("READY....\n");
fflush(stdout);
fflush(stderr);
/* Wait data for next utterance */
while ((k = cont_ad_read(cont, adbuf, 4096)) == 0){
sleep_msec(100);
}
if (k < 0)
E_FATAL("Failed to read audio\n");
/*
* Non-zero amount of data received; start recognition of new utterance.
* NULL argument to uttproc_begin_utt => automatic generation of utterance-id.
*/
if (ps_start_utt(ps, NULL) < 0)
E_FATAL("Failed to start utterance\n");
ps_process_raw(ps, adbuf, k, FALSE, FALSE);
printf("Listening...\n");
fflush(stdout);
/* Note timestamp for this first block of data */
ts = cont->read_ts;
/* Decode utterance until end (marked by a "long" silence, >1sec) */
for(;;){
//while(sleep(2)){
/* Read non-silence audio data, if any, from continuous listening module */
if ((k = cont_ad_read(cont, adbuf, 4096)) < 0)
E_FATAL("Failed to read audio\n");
if (k == 0){
/*
* No speech data available; check current timestamp with most recent
* speech to see if more than 1 sec elapsed. If so, end of utterance.
*/
if ((cont->read_ts - ts) > DEFAULT_SAMPLES_PER_SEC)
break;
}
else {
/* New speech data received; note current timestamp */
ts = cont->read_ts;
}
/*
* Decode whatever data was read above.
*/
rem = ps_process_raw(ps, adbuf, k, FALSE, FALSE);
/* If no work to be done, sleep a bit */
if ((rem == 0) && (k == 0))
sleep_msec(20);
}
/*
* Utterance ended; flush any accumulated, unprocessed A/D data and stop
* listening until current utterance completely decoded
*/
ad_stop_rec(ad);
while (ad_read(ad, adbuf, 4096) >= 0);
cont_ad_reset(cont);
printf("Stopped listening, please wait...\n");
fflush(stdout);
/* Finish decoding, obtain and print result */
ps_end_utt(ps);
hyp = ps_get_hyp(ps, NULL, &uttid);
fflush(stdout);
/* Resume A/D recording for next utterance */
if (ad_start_rec(ad) < 0)
E_FATAL("Failed to start recording\n");
if(hyp != NULL){
if(hyp == c.getValue("[General]", "Sleep")){
state = SLEEPING;
s.speakThis(c.getValue("[General]", "SleepPhrase"));
}else{
if(state != SLEEPING){
i.parse(hyp);
}
//Hack for play/pause/select
std::string hypStr = hyp;
if(hypStr == "PLAY ITEM" || hypStr == "PAUSE ITEM" || hypStr == "SELECT ITEM"){
state = SLEEPING;
}
}
}
}
}
}
pclose(pipe);
cont_ad_close(cont);
ad_close(ad);
}
/*
* Main utterance processing loop:
* for (;;) {
* wait for start of next utterance;
* decode utterance until silence of at least 1 sec observed;
* print utterance result;
* }
*/
static void
recognize_from_microphone(int outfd)
{
int16 adbuf[4096];
int32 k, ts, rem;
char const *hyp;
char const *uttid;
char word[4096];
if ((ad = ad_open_dev(NULL,
(int)cmd_ln_float32_r(config, "-adcdev"))) == NULL)
E_FATAL("Failed to open audio device\n");
/* Initialize continuous listening module */
if ((cont = cont_ad_init(ad, ad_read)) == NULL)
E_FATAL("Failed to initialize voice activity detection\n");
if (ad_start_rec(ad) < 0)
E_FATAL("Failed to start recording\n");
if (cont_ad_calib(cont) < 0)
E_FATAL("Failed to calibrate voice activity detection\n");
for (;;) {
/* Indicate listening for next utterance */
write(outfd, "READY\n", 6);
/* Wait data for next utterance */
while ((k = cont_ad_read(cont, adbuf, 4096)) == 0)
sleep_msec(100);
if (k < 0)
E_FATAL("Failed to read audio\n");
/*
* Non-zero amount of data received; start recognition of new utterance.
* NULL argument to uttproc_begin_utt => automatic generation of utterance-id.
*/
if (ps_start_utt(ps, NULL) < 0)
E_FATAL("Failed to start utterance\n");
ps_process_raw(ps, adbuf, k, FALSE, FALSE);
write(outfd, "Listening\n", 10);
/* Note timestamp for this first block of data */
ts = cont->read_ts;
/* Decode utterance until end (marked by a "long" silence, >1sec) */
for (;;) {
/* Read non-silence audio data, if any, from continuous listening module */
if ((k = cont_ad_read(cont, adbuf, 4096)) < 0)
E_FATAL("Failed to read audio\n");
if (k == 0) {
/*
* No speech data available; check current timestamp with most recent
* speech to see if more than 1 sec elapsed. If so, end of utterance.
*/
if ((cont->read_ts - ts) > DEFAULT_SAMPLES_PER_SEC)
break;
}
else {
/* New speech data received; note current timestamp */
ts = cont->read_ts;
}
/*
* Decode whatever data was read above.
*/
rem = ps_process_raw(ps, adbuf, k, FALSE, FALSE);
/* If no work to be done, sleep a bit */
if ((rem == 0) && (k == 0))
sleep_msec(20);
}
/*
* Utterance ended; flush any accumulated, unprocessed A/D data and stop
* listening until current utterance completely decoded
*/
ad_stop_rec(ad);
while (ad_read(ad, adbuf, 4096) >= 0);
cont_ad_reset(cont);
/* Finish decoding, obtain and print result */
ps_end_utt(ps);
hyp = ps_get_hyp(ps, NULL, &uttid);
snprintf(word, sizeof(word)-1, "%s: %s\n", uttid, hyp);
write(outfd, word, strlen(word));
/* Resume A/D recording for next utterance */
if (ad_start_rec(ad) < 0)
E_FATAL("Failed to start recording\n");
}
cont_ad_close(cont);
ad_close(ad);
fprintf (stderr, "listen thread exiting\n");
}
static void
utt_livepretend(void *data, utt_res_t * ur, int32 sf, int32 ef,
char *uttid)
{
char fullrawfn[FILENAME_LENGTH];
char *hypstr;
cont_ad_t *cont_ad;
ad_rec_t bogus_ad;
short samples[SAMPLE_BUFFER_LENGTH];
float32 **frames;
kb_t *kb;
int nread, n_frames, seg_n_frames;
int ts, listening;
kb = (kb_t *) data;
/* report_utt_res(ur); */
sprintf(fullrawfn, "%s/%s%s", rawdirfn, ur->uttfile, decoder.rawext);
if ((rawfd = fopen(fullrawfn, "rb")) == NULL) {
E_FATAL("Cannnot open raw file %s.\n", fullrawfn);
}
if (ur->lmname != NULL)
srch_set_lm((srch_t *) kb->srch, ur->lmname);
if (ur->regmatname != NULL)
kb_setmllr(ur->regmatname, ur->cb2mllrname, kb);
bogus_ad.sps = (int32) cmd_ln_float32_r(kb->kbcore->config, "-samprate");
if ((cont_ad = cont_ad_init(&bogus_ad, ad_file_read)) == NULL) {
E_FATAL("Failed to initialize energy-based endpointer");
}
listening = 0;
ts = 0;
seg_n_frames = 0;
while ((nread = cont_ad_read(cont_ad, samples, SAMPLE_BUFFER_LENGTH))
>= 0) {
if (nread) {
ts = cont_ad->read_ts;
if (!listening) {
char uttid[FILENAME_LENGTH];
sprintf(uttid, "%s_%.3f", ur->uttfile,
(double) ts / bogus_ad.sps);
if (s3_decode_begin_utt(&decoder, uttid) != S3_DECODE_SUCCESS)
E_FATAL("Cannot begin utterance decoding.\n");
listening = 1;
}
ptmr_start(&(st->tm));
fe_process_utt(fe, samples, nread, &frames, &n_frames);
seg_n_frames += n_frames;
if (frames != NULL) {
s3_decode_process(&decoder, frames, n_frames);
ckd_free_2d((void **)frames);
}
ptmr_stop(&(st->tm));
if (s3_decode_hypothesis(&decoder, NULL, &hypstr, NULL) ==
S3_DECODE_SUCCESS) {
if (decoder.phypdump) {
E_INFO("PARTIAL_HYP: %s\n", hypstr);
}
}
/* If the segment is too long, break it. */
if (seg_n_frames > 15000) {
s3_decode_end_utt(&decoder);
listening = 0;
}
}
else {
if (listening && cont_ad->read_ts - ts > 8000) { /* HACK */
s3_decode_end_utt(&decoder);
listening = 0;
}
}
}
fclose(rawfd);
cont_ad_close(cont_ad);
if (listening)
s3_decode_end_utt(&decoder);
}
/*
* Main utterance processing loop:
* for (;;) {
* wait for start of next utterance;
* decode utterance until silence of at least 1 sec observed;
* print utterance result;
* }
*/
static void
utterance_loop()
{
int16 adbuf[4096];
int32 k, ts, rem, score;
char const *hyp;
char const *uttid;
cont_ad_t *cont;
char word[256];
/* Initialize continuous listening module */
if ((cont = cont_ad_init(ad, ad_read)) == NULL)
E_FATAL("cont_ad_init failed\n");
if (ad_start_rec(ad) < 0)
E_FATAL("ad_start_rec failed\n");
if (cont_ad_calib(cont) < 0)
E_FATAL("cont_ad_calib failed\n");
for (;;) {
/* Indicate listening for next utterance */
printf("READY....\n");
fflush(stdout);
fflush(stderr);
/* Await data for next utterance */
while ((k = cont_ad_read(cont, adbuf, 4096)) == 0)
sleep_msec(100);
if (k < 0)
E_FATAL("cont_ad_read failed\n");
/*
* Non-zero amount of data received; start recognition of new utterance.
* NULL argument to uttproc_begin_utt => automatic generation of utterance-id.
*/
if (ps_start_utt(ps, NULL) < 0)
E_FATAL("ps_start_utt() failed\n");
ps_process_raw(ps, adbuf, k, FALSE, FALSE);
printf("Listening...\n");
fflush(stdout);
/* Note timestamp for this first block of data */
ts = cont->read_ts;
/* Decode utterance until end (marked by a "long" silence, >1sec) */
for (;;) {
/* Read non-silence audio data, if any, from continuous listening module */
if ((k = cont_ad_read(cont, adbuf, 4096)) < 0)
E_FATAL("cont_ad_read failed\n");
if (k == 0) {
/*
* No speech data available; check current timestamp with most recent
* speech to see if more than 1 sec elapsed. If so, end of utterance.
*/
if ((cont->read_ts - ts) > DEFAULT_SAMPLES_PER_SEC)
break;
}
else {
/* New speech data received; note current timestamp */
ts = cont->read_ts;
}
/*
* Decode whatever data was read above.
*/
rem = ps_process_raw(ps, adbuf, k, FALSE, FALSE);
/* If no work to be done, sleep a bit */
if ((rem == 0) && (k == 0))
sleep_msec(20);
}
/*
* Utterance ended; flush any accumulated, unprocessed A/D data and stop
* listening until current utterance completely decoded
*/
ad_stop_rec(ad);
while (ad_read(ad, adbuf, 4096) >= 0);
cont_ad_reset(cont);
printf("Stopped listening, please wait...\n");
fflush(stdout);
/* Finish decoding, obtain and print result */
ps_end_utt(ps);
hyp = ps_get_hyp(ps, &score, &uttid);
printf("%s: %s (%d)\n", uttid, hyp, score);
fflush(stdout);
/* Exit if the first word spoken was GOODBYE */
if (hyp) {
sscanf(hyp, "%s", word);
if (strcmp(word, "goodbye") == 0)
break;
}
/* Resume A/D recording for next utterance */
if (ad_start_rec(ad) < 0)
E_FATAL("ad_start_rec failed\n");
}
cont_ad_close(cont);
}
int
main(int argc, char *argv[])
{
cont_ad_t *cont;
FILE *infp;
int16 buf[512];
int listening;
int k, n_calib_samp;
int16 *calib, *cptr;
TEST_ASSERT(infp = fopen(TESTDATADIR "/chan3.raw", "rb"));
TEST_ASSERT(cont = cont_ad_init(NULL, NULL));
n_calib_samp = cont_ad_calib_size(cont);
calib = ckd_malloc(n_calib_samp * 2);
printf("Reading %d calibration samples\n", n_calib_samp);
TEST_ASSERT(fread(calib, 2, n_calib_samp, infp) == n_calib_samp);
printf("Calibrating...\n");
TEST_EQUAL(0, cont_ad_calib_loop(cont, calib, n_calib_samp));
printf("Calibrated!\n");
listening = FALSE;
cptr = calib;
while (1) {
/* Use up the calibration samples first. */
if (n_calib_samp) {
k = n_calib_samp;
if (k > 512)
k = 512;
memcpy(buf, cptr, 512 * 2);
cptr += k;
n_calib_samp -= k;
if (k < 512)
k = fread(buf + k, 2, 512-k, infp);
}
else {
k = fread(buf, 2, 512, infp);
}
/* End of file. */
if (k < 256) { /* FIXME: It should do something useful with fewer samples. */
if (listening) {
printf("End of file at %.3f seconds\n",
(double)(cont->read_ts - k) / 16000);
}
break;
}
k = cont_ad_read(cont, buf, k);
if (cont->state == CONT_AD_STATE_SIL) {
/* Has there been enough silence to cut the utterance? */
if (listening && cont->seglen > 8000) {
printf("End of utterance at %.3f seconds\n",
(double)(cont->read_ts - k - cont->seglen) / 16000);
listening = FALSE;
}
}
else {
if (!listening) {
printf("Start of utterance at %.3f seconds\n",
(double)(cont->read_ts - k) / 16000);
listening = TRUE;
}
}
}
ckd_free(calib);
cont_ad_close(cont);
fclose(infp);
return 0;
}
static void utterance_loop(int argc,char *argv[])
{
int16 adbuf[4096];
int32 k, fr, ts, rem;
char *hyp;
cont_ad_t *cont;
int count=0,i=1,flag;
char word[256];
/* Initialize continuous listening module */
if ((cont = cont_ad_init (ad, ad_read)) == NULL)
E_FATAL("cont_ad_init failed\n");
if (ad_start_rec (ad) < 0)
E_FATAL("ad_start_rec failed\n");
if (cont_ad_calib (cont) < 0)
E_FATAL("cont_ad_calib failed\n");
for (;;)
{
/* Indicate listening for next utterance */
printf ("READY....\n");
fflush (stdout); fflush (stderr);
/* Await data for next utterance */
while ((k = cont_ad_read (cont, adbuf, 4096)) == 0)
sleep_msec(200);
if (k < 0)
E_FATAL("cont_ad_read failed\n");
/*
* Non-zero amount of data received; start recognition of new utterance.
* NULL argument to uttproc_begin_utt => automatic generation of utterance-id.
*/
if (uttproc_begin_utt (NULL) < 0)
E_FATAL("uttproc_begin_utt() failed\n");
uttproc_rawdata (adbuf, k, 0);
printf ("Listening...\n"); fflush (stdout);
/* Note timestamp for this first block of data */
ts = cont->read_ts;
/* Decode utterance until end (marked by a "long" silence, >1sec) */
for (;;) {
/* Read non-silence audio data, if any, from continuous listening module */
if ((k = cont_ad_read (cont, adbuf, 4096)) < 0)
E_FATAL("cont_ad_read failed\n");
if (k == 0) {
/*
* No speech data available; check current timestamp with most recent
* speech to see if more than 1 sec elapsed. If so, end of utterance.
*/
if ((cont->read_ts - ts) > DEFAULT_SAMPLES_PER_SEC)
break;
} else {
/* New speech data received; note current timestamp */
ts = cont->read_ts;
}
/*
* Decode whatever data was read above. NOTE: Non-blocking mode!!
* rem = #frames remaining to be decoded upon return from the function.
*/
rem = uttproc_rawdata (adbuf, k, 0);
/* If no work to be done, sleep a bit */
if ((rem == 0) && (k == 0))
sleep_msec (20);
}
/*
* Utterance ended; flush any accumulated, unprocessed A/D data and stop
* listening until current utterance completely decoded
*/
ad_stop_rec (ad);
while (ad_read (ad, adbuf, 4096) >= 0);
cont_ad_reset (cont);
printf ("Stopped listening, please wait...\n"); fflush (stdout);
#if 0
/* Power histogram dump (FYI) */
cont_ad_powhist_dump (stdout, cont);
#endif
/* Finish decoding, obtain and print result */
uttproc_end_utt ();
if (uttproc_result (&fr, &hyp, 1) < 0)
E_FATAL("uttproc_result failed\n");
/*obtaining the results*/
sscanf (hyp, "%s", word);
printf ("%d: %s\n", fr,word); fflush (stdout);
win=wnck_screen_get_active_window(scr);
tmp=wnck_screen_get_windows(wnck_window_get_screen(win));
i=g_list_index(tmp,win);/*the place value of window in the list*/
printf("<<<<<<<<<<<<<<>>>>>>>>>>>>>>i:%d,%s\n\n\n",i,wnck_window_get_name(win));
count=0;
while(tmp!=NULL)
{
printf("%d:%s\n\n",count,wnck_window_get_name(tmp->data));
tmp=tmp->next;count++;
}
/*comparison and action for DAKKU */
if(strcmp(word,"PADU")==0)
g_spawn_command_line_async("totem --play",NULL);
if(strcmp(word,"EMACS")==0)
g_spawn_command_line_async("emacs",NULL);
if(strcmp(word,"SAMAYAM")==0)
g_spawn_command_line_async(DHVANISCRIPT,NULL);
if(strcmp(word,"VALAPARATHU")==0)
g_spawn_command_line_async("epiphany",NULL);
if(strcmp(word,"EAZHUTHIDAM")==0)
g_spawn_command_line_async("gedit",NULL);
/*Minimizing current active window*/
if (strcmp (word, "CHURUKKU") == 0)
wnck_window_minimize(win);
/*Moving focus(active window)towards the left of the panel.The active window is changed to
the next normal window on the left.effect of alt+tab key press*/
if(strcmp(word,"ADUTHATHU")==0)
{
win=wnck_screen_get_active_window(scr);
tmp=wnck_screen_get_windows(wnck_window_get_screen(win));
while(tmp!=NULL)/*while traces the current active window through the list*/
{
printf("tracing:current:%s\n\ntmp:%s\n\n",wnck_window_get_name(win),wnck_window_get_name(tmp->data));
if(tmp->data==win)
{
printf("BREAKED with tmp:%s\n",wnck_window_get_name(tmp->data));
break;
}
tmp=tmp->next;
}
if(tmp==NULL){printf("BULL SHIT GIVE A WINDOW IN THE LIST\n\n");}//exit(1);}
if(tmp->next==NULL)/*shifting back to the first window by refreshing the list*/
tmp=wnck_screen_get_windows(wnck_window_get_screen(win));
else
tmp=tmp->next;
printf("cuow:%s\n\n",wnck_window_get_name(tmp->data));
while(tmp!=NULL)
{
printf("tmp in while:%s\n\n",wnck_window_get_name(tmp->data));
if(wnck_window_get_window_type(tmp->data)==WNCK_WINDOW_NORMAL)
{
wnck_window_activate(tmp->data,0);
flag=1;
break;
}
else
tmp=tmp->next;
}
if(flag==0)
{
printf("FLAG==0 DETECTED:\n");
tmp=wnck_screen_get_windows(wnck_window_get_screen(win));
while(tmp!=NULL)
{
printf("tmp in last while:%s\n",wnck_window_get_name(tmp->data));
if(wnck_window_get_window_type(tmp->data)==WNCK_WINDOW_NORMAL)
{
wnck_window_activate(tmp->data,0);
break;
}
else
tmp=tmp->next;
}
}
}
if(strcmp(word,"VALUTHAKKU")==0)
{
if(wnck_window_get_window_type(win)!= WNCK_WINDOW_NORMAL)
{
if(wnck_window_get_window_type(wnck_screen_get_previously_active_window(scr))==WNCK_WINDOW_NORMAL)
win=wnck_screen_get_previously_active_window(scr);
}
wnck_window_unminimize(win,0);
}
while(gtk_events_pending())/*gtk probing and refreshing the win and tmp*/
{
gtk_main_iteration();
win=wnck_screen_get_active_window(scr);
tmp=wnck_screen_get_windows(wnck_window_get_screen(win));
}
/* Resume A/D recording for next utterance */
if (ad_start_rec (ad) < 0)
E_FATAL("ad_start_rec failed\n");
}
cont_ad_close (cont);
}
