intermediate results of the decoding

[Gyozo Karsai]

Dear 'PocketSphinx' team,
I would like to transcribe some speech in wav file into text with the help of PocketSphinx. But if the sound file is too long then the result of decoding appears only after having recognized the entire file, that will take a long time. Is it possible to display the intermediate results of the decoding as to have outcomes while the process.

[Nickolay V. Shmyrev]

Sure, it is perfectly possible, it depends a how exactly do you use the library though. Is it Python API or something else.

[Gyozo Karsai]

Dear 'PocketSphinx' team,
I use a simple visual C++ code for driving of PocketSphinx. I embed the code of PocketSphinx into visual C++ project and I call the appropriate (I hope) methods for the processing any sound file. There is the current code sniplet:
int16 adbuf[2048];
ps_start_utt(ps);
utt_started = FALSE;

   rawfd = fopen(fname, "rb");  //processed input sound file
    while((k = fread(adbuf, sizeof(int16), 2048, rawfd)) > 0) {
        ps_process_raw(ps, adbuf, k, FALSE, FALSE);
        in_speech = ps_get_in_speech(ps);
        if(in_speech && !utt_started) {
            utt_started = TRUE;
        }
        if(!in_speech && utt_started) {
            ps_end_utt(ps);
            hyp = ps_get_hyp(ps, NULL);
            if(hyp != NULL) {
              //printf("%s\n", hyp);
                canalToOut(hyp);  //writing of result of decoding
            }

            if(print_times) print_word_times();
            fflush(stdout);

            ps_start_utt(ps);
            utt_started = FALSE;
        }
    }
    ps_end_utt(ps);

[Nickolay V. Shmyrev]

You can use ps_get_hyp after process_raw to get intermediate result.

ps_process_raw(ps, adbuf, k, FALSE, FALSE);
hyp = ps_get_hyp(ps, NULL); // <-- You can get intermediate hypothesis here
in_speech = ps_get_in_speech(ps);
if(in_speech && !utt_started) {
    utt_started = TRUE;
}

[Gyozo Karsai]

Dear 'PocketSphinx' team,
thanks very much for rapid answer