sphinx3-doc 0.8-0ubuntu1 / usr / share / doc / sphinx3 / s3_overview.html

This file is indexed.

/usr/share/doc/sphinx3/s3_overview.html is in sphinx3-doc 0.8-0ubuntu1.

This file is owned by root:root, with mode 0o644.

The actual contents of the file can be viewed below. 

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
<!DOCTYPE HTML PUBLIC "-//w3c//DTD HTML 4.01//EN">
<html>
<head>
  <title>Sphinx Decoders</title>
                 
  <style type="text/css">
      body { margin-left: 6%; margin-right: 3%; background: white; }
      h1 { color: green; align: center; }
      h2 { margin-left: -6%; color: green; margin-top: 1em; }
      h3 { margin-left: -3%; color: green; margin-top: 1em; }
      h4,h5,h6 { color: green; }
      pre { margin-left: 3%; font-family: monospace; }
      code { font-family: monospace; font-weight: bold; }
      legacy { color: gray; }
      div.wheatbox { background: wheat; padding: 0.5em; border: solid; border-width: thin; }
      div.silverbox { background: silver; padding: 0.5em; border: solid; border-width: thin; }
      div.endsec { background: silver; border: solid; border-width: thin; }
    </style>

<SCRIPT  TYPE="text/javascript">

<!--
// protected email script by Joe Maller
// JavaScripts available at http://www.joemaller.com
// this script is free to use and distribute
// but please credit me and/or link to my site

function printAddress(userid, name){
emailE='cs.cmu.edu'
emailE=(userid + '+sourceforge@' + emailE)
document.write('<A href="mailto:' + emailE + '">' + name + '</a>')
}
 //-->
</script>

<NOSCRIPT>
    <em>Email address protected by JavaScript.<BR>
    Please enable JavaScript to contact me.</em>
</NOSCRIPT>

</head>
    <body>
       
<h1><center><u>Sphinx Decoders</u></center></h1>
       
<center>       Mosur K. Ravishankar (<em>aka</em> Ravi Mosur)<br>
        Sphinx Speech Group<br>
        School of Computer Science<br>
        Carnegie Mellon University<br>
        Pittsburgh, PA 15213<br>
</center>
                 
<h2><u>What's Around and Where</u></h2>

<p>In the table below, you will find information about the several
versions of the Sphinx decoder. Descriptions in <legacy>gray</legacy>
are older versions, and are not being maintained. Follow the links to
find instructions on how to download the packages.
</p>
 
<table border="1" cellpadding="7">
        <tbody>
     <tr>
       <th>Name</th>
       <th>Remarks</th>
     </tr>
	<tr>
	<td><a href="./s3_description.html">Sphinx 3.6 (Re-architected Version)</a></td> 
	<td> 
	<ul>
  	    <li>Location: <a href="http://cmusphinx.org">Open source</a>, release available.</li>
  	    <li>Based on Sphinx-3.5. </li>

	    <li> Here are several new features in Sphinx 3.6 </li> 
	    <ul> 
		<li> Further speed-up of CIGMMS in the 4 level GMM Computation Schemes (4LGC) </li>
		<li> Multiple regression classes an MAP adaptation in SphinxTrain </li> 
		<li> Better support in using LM in Sphinx 3.X </li> 
		<li> FSG search is now supported. This is adapted from Sphinx 2. </li> 
		<li> Support of full triphone search in flat lexicon search </li> 
		<li> Some support of different character sets other than of 
        Sphinx 3.X.  Models in multiple languages are now tested in 
        Sphinx 3.X. </li> 
	    </ul> 
	</ul> 
	 </td> 
	</tr> 
         <tr>
       <td><legacy>Sphinx-3.5 (Live-mode APIs, Speaker Adaptation and Code Convergence)</legacy></td>
  	<td> 	         
      <ul>
  	    <li>Location: <a href="http://cmusphinx.org">Open source</a>, release available.</li>
  	    <li>Based on Sphinx-3.4 and Sphinx 3.0. </li>
	    <li>Live-mode APIs are implemented an thoroughly tested </li>
	    <li>Speaker adaptation based on single regression class
	    is implemented. </li>
	    <li> Four tools of s3.0 are now incorporated in Sphinx 3.x 
	    <ul>
  		<li><code>align (s3align in 3.0)</code>: Forced alignment</li>
  		<li><code>allphone (s3allphone in 3.0)</code>: Allphone decoding</li>
  		<li><code>astar (s3astar in 3.0)</code>: A* search, nbest generation</li>
  		<li><code>dag (s3dag in 3.0)</code>: Shortest-path search</li>
	    </ul>
	    </li>
	    <li> The feature extraction libraries of sphinx 3 is now EXACTLY the same as SphinxTrain</li>
	    <li> Corresponding changes in SphinxTrian </li>
	    <li> Two new tools are introduced: 
<ul>
	    <li><code> mllr_solve </code>: compute the regression matrix based on MLLR algorithm. </li>
	    <li><code> mllr_transform </code>: given a regression matrix, this program converts the mean based on linear transformation.</li>
	    </ul> 
	    </li> 
	    <li> The command line interface of all SphinxTrain's tools are now unified. 
	    </li>
      </ul>
  	</td>
        </tr>
         <tr>
       <td><legacy>Sphinx-3.4 (Fast GMM computation)</legacy></td>
  	<td> 	         
      <ul>
  	    <li>Location: <a href="http://cmusphinx.org">Open source</a>, module <b>archive_s3/s3.4</b> in cvs tree.</li>
  	    <li>Based on Sphinx-3.3, Fast GMM computation is implemented
          <ul>
  		<li>frame down-sampling</li>
  		<li>CI-based GMM selection</li>
  		<li>VQ-based and SVQ-based Gaussian Selection</li>
  	        <li>Support of SVQ with arbitrary number of sub-vectors.</li>
          </ul>
  	    </li>
  	    <li> Phoneme look-ahead </li>
	    <li> Support class-based LM and multiple LM </li>
	    </li>
      </ul>
  	</td>
        </tr>

         <tr>
       <td><legacy>Sphinx-3.3 (fast decoder)</legacy></td>
  	<td>
      <ul>
  	    <li>Location: <a href="http://cmusphinx.org">Open source</a>, module <b>archive_s3/s3.3</b> in cvs tree.</li>
  	    <li>Fast Sphinx-3 decoder using lextree organization: 	          
      
          <ul>
  		<li>5-10x real time speed on large vocabulary tasks (measured at 1999)</li>
  		<li>Continuous density acoustic models only</li>
  		<li>Batch-Mode or live operation</li>
  	                 
          </ul>
  	    </li>
<li>Other tools
<ul>
  	    <li><code>gausubvq</code>: Sub-vector clustered acoustic model building, needed for fast acoustic model evaluation</li>
</ul>
  	    </li>
  	         
      </ul>
  	</td>
        </tr>
               <tr>
         <tr>
       <td><legacy>Sphinx-3.2</legacy></td>
  	<td> 	         
      <ul>
  	    <li>Location: <a href="http://cmusphinx.org">Open source</a>, module <b>archive_s3/s3.2</b> in cvs tree.</li>
  	    <li>Same features as s3.3, but capable of batch-mode operation only.</li>
      </ul>
  	</td>
        </tr>
         <tr>
       <td>Sphinx-3 (slow decoder)</td>
  	<td> 	         
      <ul>
  	    <li>Location: <a href="http://cmusphinx.org">Open source</a>, module <b>archive_s3/s3</b> in cvs tree.</li>
  	    <li>Original Sphinx-3 decoder</li>
  	    <li>Slow; 50-100x real time speed on large vocabulary tasks (measured at 1999)</li>
  	    <li>Any kind of acoustic model (discrete, semi-continuous, continuous, 
others)</li>
  	    <li>Major applications: 	                 
          <ul>
  		<li><code>s3decode</code> and <code>s3decode-anytopo</code>: Speech-to-text 
Decoding</li>
  		<li><code>s3align</code>: Forced alignment</li>
  		<li><code>s3allphone</code>: Allphone decoding</li>
  		<li><code>s3astar</code>: A* search, nbest generation</li>
  		<li><code>s3dag</code>: Shortest-path search</li>
  	                 
          </ul>
  	    </li>
  	    <li>Other utilities: 	                 
          <ul>
  		<li><code>stseg-read</code>: State-segmentation binary file reader</li>
  		<li><code>sen2s2</code>: Sphinx-II "sendump" file creation from Sphinx-3
 		  acoustic model</li>
  	                 
          </ul>
  	    </li>
      </ul>
  	</td>
        </tr>
         <tr>
       <td>Sphinx-2 (fbs8)</td>
  	<td> 	         
      <ul>
  	    <li>Location: <a href="http://cmusphinx.org">Open source</a>, release available.</li>
  	    <li>Sphinx-II decoder</li>
  	    <li>Real-time operation</li>
  	    <li>Semi-continuous, Sphinx-II acoustic models only (Sphinx-II format)</li>
  	    <li>User applications support: 	                 
          <ul>
  		<li>Compiled into a library with a straightforward API for building 		 
 speech-enabled applications</li>
  		<li>Continuous-listening support</li>
  		<li>Dynamic language model loading and switching</li>
  	                 
          </ul>
  	    </li>
  	    <li>Several test applications: 	                 
          <ul>
  		<li>Basic dictation with and without "push-to-talk"</li>
  		<li>Basic audio recording and playback</li>
  		<li>Audio segmentation using the continuous listener</li>
  	                 
          </ul>
  	    </li>
  	    <li>Additional recognition modes: 	                 
          <ul>
  		<li>Forced alignment</li>
  		<li>Allphone decoding</li>
  		<li>A* search, nbest generation</li>
  		<li>Shortest-path search</li>
  	                 
          </ul>
  	    </li>
  	         
      </ul>
  	</td>
        </tr>
        
  </tbody> 
</table>
              
<div class="endsec">&nbsp;</div>
                 
<address>Maintained by <SCRIPT>printAddress("egouvea", "Evandro B. Gouv&ecirc;a")</script> and  <script>printAddress("archan", "Arthur Chan")</script></address>
<!-- Created: Fri Mar  3 11:03:47 EST 2000 -->
<!-- hhmts start -->
Last modified: Wed Jul 26 13:28:16 Eastern Daylight Time 2006
<!-- hhmts end --> 
</body>
</html>

APT Browse - Built by Thomas Orozco.