Human Emulator wiki

recognize_file

recognize_file($path,$dict,$jsgf,$hmm,$add_params="",$timeout=60); - recognize an audio file as speech using the Open Source Speech Recognition Toolkit CMU Sphinx (Available from version 4.10.7)
- the file must have a resolution of 128 Kb/s (8 kHz) ( you can Change the file format via $sound->convert_file )

The function accepts parameters as input:

$path – path to the audio file to be recognized (wav format, sampling must meet the sampling requirements of the model used)

$dict – dictionary path (the path must be relative to the Audio Tools folder)

$jsgf – path to the gramatics file (format JSGF (the path must be relative to the Audio Tools folder)

$hmm – directory path hmm models (the path must be relative to the Audio Tools folder)

$add_params – additional command-line parameters to run pocketsphinx_continuous.exe

$timeout – maximum time allowed for this operation

After testing, the function returns the result of its work to the script :

true – a sequence of numbers dictated in the audio file (separated by a space)

false – the operation was not completed successfully

PHP example:

<?php $xhe_host = "127.0.0.1:7010";
 
//connect the object to control the emulator, if not already connected
if(!isset($path))
   $path = "../../../Templates/xweb_human_emulator.php";
require($path);
 
//Start
echo "<hr> <font color = blue> sound->".basename(__FILE__)."</font> <hr>";
 
//1
echo "1.Prepare the file:";
echo $sound->convert_file("test/mp3/4196092774.mp3", "test/mp3/4196092774.wav", "8000")." n";
 
//2
echo "2.Recognize the file with English numbers(0-9) using CMU Sphinx:";
echo $sound->recognize_file("test/mp3/4196092774.wav", "digits \ digits.dict", "digits \ digits.gram", "cmusphinx-en-us-8khz-5.2", "- samprate 8000 -remove_dc yes -remove_noise no -vad_threshold 3.4 -vad_prespeech 19 -vad_postspeech 37 -silprob 2.5 ")."  N ";
 
//end
echo "<hr> <br>";
 
//Quit
$app->quit();
?>

Python example:

# Additional paths
import sys
sys.path.insert(0, '../../../Templates PY/')
 
xhe_host = "127.0.0.1:7010"
from xweb_human_emulator import *
 
# Start
echo("<hr> <font color = blue> sound.xxxxxxxxx </font> <hr>")
 
# 1
echo("1.Prepare the file:")
echo(sound.convert_file("test/mp3/4196092774.mp3", "test/mp3/4196092774.wav", "8000"), "
")
 
# 2
echo("2.Recognize file with English numbers(0-9) using CMU Sphinx:")
echo(sound.recognize_file("test/mp3/4196092774.wav", "digits \ digits.dict", "digits \ digits.gram", "cmusphinx-en-us-8khz-5.2", "- samprate 8000 -remove_dc yes -remove_noise no -vad_threshold 3.4 -vad_prespeech 19 -vad_postspeech 37 -silprob 2.5 "),"
")
 
# end
echo("<hr> <br>")
 
# Quit
app.quit()

C# example:

#region using
 
using System;
using System.Diagnostics;
using System.Collections.Generic;
using System.Linq;
using System.IO;
using System.Text;
using System.Threading;
 
using XHE;
using XHE.XHE_DOM;
using XHE.XHE_System;
using XHE.XHE_Window;
using XHE.XHE_Web;
 
#endregion
 
 class Program: XHEScript
 {
static void Main(string [] args)
{
//init XHE
server = "127.0.0.1:7021";
InitXHE();
 
//Start
echo("<hr> <font color = blue> sound.recognize_file </font> <hr>");
 
//1
echo("1.Prepare the file:");
echo(sound.convert_file("test/mp3/4196092774.mp3", "test/mp3/4196092774.wav", "8000") + " n");
 
//2
echo("2.Recognize the file with English numbers(0-9) using CMU Sphinx:");
echo(sound.recognize_file("test/mp3/4196092774.wav", "digits \ digits.dict", "digits \ digits.gram", "cmusphinx-en-us-8khz-5.2", "- samprate 8000 -remove_dc yes -remove_noise no -vad_threshold 3.4 -vad_prespeech 19 -vad_postspeech 37 -silprob 2.5 ") +"  n ");
 
//end
echo("<hr> <br>");
 
app.quit();
}
}

Java Script example:

//connect the object to control the emulator, if not already connected
xhe_host = "127.0.0.1:7022";
echo = require("../../../Templates JS/xweb_human_emulator.js");
 
//Start
echo("<hr> <font color = blue> sound.play_sound </font> <hr>");
 
//1
echo("1.Prepare the file:");
echo(sound.convert_file("test/mp3/4196092774.mp3", "test/mp3/4196092774.wav", "8000") + " n");
 
//2
echo("2.Recognize the file with English numbers(0-9) using CMU Sphinx:");
echo(sound.recognize_file("test/mp3/4196092774.wav", "digits \ digits.dict", "digits \ digits.gram", "cmusphinx-en-us-8khz-5.2", "- samprate 8000 -remove_dc yes -remove_noise no -vad_threshold 3.4 -vad_prespeech 19 -vad_postspeech 37 -silprob 2.5 ") +"  n ");
 
//end
echo("<hr> <br>");
 
//Quit
app.quit();

Human Emulator wiki

User Tools

Site Tools

Sidebar

recognize_file

Page Tools