Speech Recognition HOWTO Stephen Cook scook@gear21.com - ú{êó htakashi@yabumi.com Revision History Revision v1.2 February 5, 2002 Added more commercial software listings (sent by Mayur Patel). Revision v1.1 October 5, 2001 Revised by: scc Added info for Vocalis Speechware. Fixed/Updated various other items. Revision v1.0 November 20, 2000 Revised by: scc Added info on L and H and HTK Revision v0.5 September 13, 2000 Revised by: scc Initial HOWTO Submission Linux ãÅÌ©®¹ºF¯ (ASR) ªÈPÉÈè èÜ·. JÒ¾¯ÅÈ [UÅàüèÂ\ÈàÌà èÜ·. ±Ì¶ÅÍ, ¹ºF¯ÌîbÆ» êçüèÂ\È\tgEFAÉ¢ÄLqµÜ·. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª Table of Contents 1. @IÈÓ 1.1. ì /CZX 1.2. ÆÓ 1.3. ¤W 2. Ou« 2.1. 2.2. Ó« 2.3. Rg/ÅVîñ/tB[hobN 2.4. ToDo 2.5. üùð 3. ͶßÉ 3.1. ¹ºF¯Ìîb 3.2. ¹ºF¯Ì^Cv 3.3. p@Æp 4. n[hEFA 4.1. TEhJ[h 4.2. }CN 4.3. Rs [^/vZbT 5. ¹ºF¯\tgEFA 5.1. t[\tgEFA 5.1.1. XVoice 5.1.2. CVoiceControl/kVoiceControl 5.1.3. Open Mind Speech 5.1.4. GVoice 5.1.5. ISIP 5.1.6. CMU Sphinx 5.1.7. Ears 5.1.8. NICO ANN Toolkit 5.1.9. Myers' Hidden Markov Model Software 5.1.10. Jialong He's Speech Recognition Research Tool 5.1.11. ܾ¼Éà èÜ·©? 5.2. ¤p\tgEFA 5.2.1. IBM ViaVoice 5.2.2. Vocalis Speechware 5.2.3. Babel Technologies 5.2.4. SpeechWorks 5.2.5. Nuance 5.2.6. Abbot/AbbotDemo 5.2.7. Entropic 5.2.8. ¼Ì¤p»i 6. ¹ºF¯Ìठ6.1. ÇÌæ¤ÉF¯µÄ¢é© 6.2. fBW^I[fBIÌîb 7. oŨ 7.1. Ð 7.2. C^[lbg 8. ú{êóÉ墀 1. @IÈÓ 1.1. ì /CZX (ó: ´¶ðcµÜ·.) This document is copyrighted (c) 2000-2002 Stephen C. Cook. LICENSE: This document may be reproduced and distributed in whole or in part, in any medium physical or electronic, provided that this license notice is displayed in the reproduction. Commercial redistribution is permitted and encouraged. Thirty days advance notice, via email to the author, of redistribution is appreciated, to give the author time to provide updated documents. CZX: ±ÌCZXª»ÌÉ\¦³êÄ¢éÀè, ±Ì¶Ì êܽÍSð, ¨I é¢ÍdqIÈ çäé}ÌÅC³µ, ¡»·é ±ÆªÅ«Ü·. ¤IÈÄzzàÂ, §³êĢܷ. 30úOàÁÄ, ì ÒÉ Email ðʶÄ, ÄzzÌÊmðêéƤ굢ŷ, ìÒÉÅV̶ ðpÓ·éÔ𺳢. All modified documents, including translations, anthologies, and partial documents, must meet the following requirements: |óâA\W[, ¶ÌêðÜßÄ, SÄÌC³³ê½¶ÍȺÌð ð«µÈ¯êÎÈèܹñ: E Modified versions must be labeled as such. C³³ê½ÅÍ»Ì|ª¦³êĢȯêÎÈèܹñ. E The person making the modifications must be identified. C³ðsÈÁ½lªÁè³êĢȯêÎÈèܹñ. E Acknowledgement of the original author must be retained. IWiÌÒ̳FªÛ½êĢȢ¯êÎÈèܹñ E The location of the original unmodified document be identified. IWiÌÏXO̶ÌêªÁè³êĢȯêÎÈèܹñ. E The original author's name(s) may not be used to assert or imply endorsement of the resulting document without the original author's permission. ´Ò̳, ´Ò̼OðgÁÄ, Ê̶ÌmFð壵½ èæµ½èµÈ¢Åº³¢. E The author be notified by email of the modification in advance of redistribution. ÄzzÌOÉ, C³É¢ÄÒÉ email ÅÊmµÄ¾³¢. E As a special exception, anthologies of LDP documents may include a single copy of these license terms in a conspicuous location within the anthology and replace other copies of this license with a reference to the single copy of the license without the document being considered "modified" for the purposes of this section. ÁÊÈáOƵÄ, LDP ̶ÌA\W[Í, ±êçÌCZXð ÌPêÌRs[ðA\W[ÌàÌÚ§ÂêÉÜÝ, ±ÌCZ X̼ÌRs[ð, »ÌPêÌCZXÌRs[ÖÌQÆÅ·¦é±Æ ª èÜ·. ±ÌêÍ{ßÌÚI©çÍÏXƩȳêܹñ. Mere aggregation of LDP documents with other documents or programs on the same media shall not cause this license to apply to those other works. ¯¶fBAãż̶âvOðWß½ LDP ¶ÌPÈéWÌÍ, »êç̼ÌìiɱÌCZXðKp·é±ÆÍ èܹñ. All translations, derivative documents, or modified documents that incorporate this document may not have more restrictive license terms than these, except that you may require distributors to make the resulting document available in source format. zzÒɶ¬¨Ì¶ð\[XÌ`®ÅüèÅ«éæ¤Éßéêð¢Ä, SÄÌ|ó, h¶µ½¶, é¢Í±Ì¶ðgÝñÅC³³ê½¶Í ±êÈãµµ¢CZXð½¹ÄÍ¢¯Ü¹ñ. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 1.2. ÆÓ (ó: ´¶ðcµÜ·.) The author disclaims all warranties with regard to this document, including all implied warranties of merchantability and fitness for a certain purpose; in no event shall the author be liable for any special, indirect or consequential damages or any damages whatsoever resulting from loss of use, data or profits, whether in an action of contract, negligence or other tortious action, arising out of or in connection with the use of this document. ÒÍ, SÄ̤sתÂ\Å é±ÆÌÃÙÌÛØ, éÚIÖK·é± ÆðÜßı̶ÉÖ·éSÄÌÛØðúüµÜ·; ÇÌæ¤Èoª ÁÄà, ±Ì¶ÌgpÆÌpªèÌàOÅN±é, KñÌÌ®, Ó é¢Í¼Ìs@s×ÉæéàÌÅ ë¤Æ, çäéÁÊÈ, ÔÚIܽÍ, ÊIȹQâgp, f[^, v̹¸Éæé¹QÈÇÉεÄìÒÍÓC ð¢Ü¹ñ. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 1.3. ¤W ±Ì¶ÉÜÜêéSÄ̤WÍ»ê¼êÌLÒÌì /o^¤WÅ·. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 2. Ou« 2.1. ±Ì¶Í¹ºF¯ÌwK»¡ª è, µÄÝæ¤ÆµÄ¢é©ç xÌ Linux [Uð^[QbgɵĢܷ. ܽ, »¡ðÁ½JÒ Ì½ßɹºF¯ÉÖ·évO~OÌîbÉ¢Äàྵܷ. ÇÌæ¤È¹ºF¯\tgEFAÆJpÌCuª Linux ÅgpÅ«é Ì©ð²×n߽ƫɱ̶ð«Í¶ßܵ½. Linux ãÅÌ©®¹º F¯ (ASR ܽÍPÉ SR) Í¿å¤Ç{ÌðöµÍ¶ß½±ÆëÅ, ±Ì¶ ųµ¢ûüÖãµÅ«é±ÆðèÁĢܷ - ASR ZpÌ[UÆJ Ò̼ûðT|[g·é±ÆÅ. ±Ì¶Í SR ÌZpÉ¢ÄÍGêĢܹñ, »ÌãèÉ "HOWTO" Æ¢¤ ¤ÊÉWµÄ¢Ü· (±êÍ HOWTO Å·©çc). ±±ÅJo[Å«Ä¢È ¢±ÆÉ¢ÄÍ, »¡ðÁ½ÇÒª{âLðT¹éæ¤ÉoŨÌßð pӵܵ½. ±êªLinux ãÌ ASR É¢ÄÌÅIIÈñÆ¢¤±ÆÅÍ èܹñ. ±Ì¶ÌÅVÅÍ, LDP ÌA[JCuð`FbN·é©, http:// www.gear21.com/speech/index.html©çüèµÄ¾³¢. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 2.2. Ó« ±Ì¶ð©¼µ, µÄ¾³Á½ÈºÌlXɴӵܷ: E Jessica Perry Hekman E Geoff Wexler ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 2.3. Rg/ÅVîñ/tB[hobN Rgâ, ñÄ, üù, ÅVîñª êÎ, ܽ, ½¾ ASR É¢Ä`b gµ½¢Æ«à, ÌAhX scook@gear21.com <mailto:scook@gear21.com> É Email 𺳢. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 2.4. ToDo Ⱥ̱ƪ "to do" ƵÄcÁĢܷ: E oŨÌßÉà¾ðÁ¦é. E oŨÌßÉæè½Ì{ðÁ¦é. E æè½ÌNðà¾t«ÅÁ¦é. E ASR VXeÌèÉ¢ÄÌà¾ð[À³¹é. E FFT ÆtB^[Ìà¾ðÁ¦é. E DSP Ì´Ìà¾ðÁ¦é. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 2.5. üùð v0.1 ÅÌÄ 2000N 8 v0.5 ÅIÄ 2000N 9 ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 3. ͶßÉ 3.1. ¹ºF¯Ìîb ¹ºF¯ÆÍ, Rs [^ ( é¢Í¼Ì^CvÌ@B) ªbµ¾tðF¯ ·éÅ·. î{IÉÍ, Rs [^ÉüÁÄbµ, »Ì¾tªRs [^ɳµF¯³êéÆ¢¤Ó¡Å·. ȺÌè`͹ºF¯ÌZpðð·é½ßÉKvÈîbÅ·. b bÍ, 1ÂÌÓ¡ð\·Pêâ¢Â©Ì¾tðRs [^Éü©ÁÄ º·é (b·) ±ÆÅ·. bÍPêÅ Á½è, ¾tÅ Á½è, ¶ Å Á½è, é¢Í¡Ì¶Å Á½èµÜ·. bÒÖÌ˶ bÒÉ˶·éVXeÍÁèÌbÒðÎÛƵÄÝv³êÜ·. »ÌV XeÍêÊÉ, »ÌÁèÌbÒ̺ÉεÄͳmÅ·ª, ¼ÌbÒ Å͸xª¸ÁÆ«ÈèÜ·. »êçÍbÒªêè̺ƬxÅb·± Æð¼èµÄ¢Ü·. bÒÉ˶µÈ¢VXeÍlXÈbÒÉü¯ÄÝ v³êÜ·. «Ì éVXeÍÊ, bÒÉ˶µÈ¢VXeÆ µÄX^[gµ, wKZpðpµÄF¯¸xðßé±ÆÅbÒÉK µÄ¢«Ü·. êb êb ( é¢Í«) ÆÍ, SR VXeÉF¯³êé½ß̾tâbÌ XgÅ·. êÊÉ, Rs [^ÉÆÁÄÍÈ¢êbÌÙ¤ªF¯µ â·, êbª½ÈéÙÇF¯ª¢ïÉÈèÜ·. ÊÌ«ÆÍÙÈ è, »ê¼êÌÚÍPêÅÍ èܹñ. »êçͶâ¶ÍÙÇ·È é±Æà èÜ·. È¢êbÍ1©2ÂÌF¯³ê½¶ (á¦Î "Wake up") µ©È¢©àµêܹñª, ÆÄརêbÅÍ 10 êÈãÆÈè Ü·. ¸x F¯uÌ\Íͻ̸xðªè·é±ÆÉæÁÄ, é¢Íܽ, b³ ê½¾tðÇêç¢F¯·é©ÉæÁIJ×é±ÆªÅ«Ü·. ±êÍ bð³mÉÁè·é¾¯ÅÈ, bªêbÉÜÜêÄ¢é©Ç¤©ð Áè·é±ÆàÜñŢܷ. Ç¢ ASR VXeÍ 98% Èã̸xª èÜ·. éVXe̸xÌeÍÍÍ»ÌprÉ˶µÜ·. wK bÒÉ·é\Íð¹ºF¯à èÜ·. VXeª±Ì\Íðà ÁÄ¢éÆ«Í, wK³¹é±ÆªÅ«Ü·. ASR VXeÍbÒÉW IȾtâêÊIȾtðJèÔ³¹, ärÌASYðÁèÌbÒ É²a·é±ÆÅwK³êÜ·. êÊÉF¯uðwK³¹é±ÆÅ, » ̸xÍüãµÜ·. wKÍ, bµûâ éíÌPê̹ª¤ÜÅ«È¢bÒÉàp³ê Ü·. bÒªñöêѵÄbðJèÔ·Àè, wK@\Ì é ASR VX eÍK·é±ÆªÂ\ŵå¤. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 3.2. ¹ºF¯Ì^Cv ¹ºF¯ÌVXeÍ, ÇÌæ¤È^CvÌbðF¯·é\ÍðÁÄ¢é ©ÉæÁÄ, ô©ÌNXɪ޷é±ÆªÅ«Ü·. ±Ìæ¤ÈNXÍ bÒª¢Âbðnß, ¢ÂI¦½Ì©ðªè·é\ͪ ASR Ìïµ³Ì 1 ŠéÆ¢¤ÀÉîâĢܷ. ½ÌpbP[WªgpÌ[hÉæ ÁÄ, ¡ÌNXÉKµÜ·. ǧµ½¾t ǧµ½¾tÌF¯ÉÍ, »ê¼êÌb²ÆÉTvEBhE (T vÌJn©çI¹ÌúÔ)ÌOãɹÌÈ¢Ô (I[fBIM̳ ¢óÔ)ªKvÆÈèÜ·. F¯uªPêðó¯æéÆ¢¤í¯ÅÈê xÉÍbÍêÂÆ¢¤Ó¡Å·. ±ÌVXeÅÍÊÈÌÅ·ª, ``¹ºüÍóÔ / F¯óÔ'' Æ¢¤ 2 ÂÌóÔª é½ßCbÒÍƬ êƬêÉb³È¯êÎÈèܹñ (ºªÆ¬ê½Æ«ÉF¯ðµÄ ¢Ü·). ǧµ½bͱÌNXÅÍæèÇ¢¼O©àµêܹñ. A±µ½¾t A±µ½¾t ( é¢Íæè³mÉ 'A±µ½b') ÌVXeÍǧµ ½¾tÌVXeÉĢܷª, ÔÉÅZÌx~ðͳÝȪç '±¯ ĺ³êé' ÂÊÌbðF¯µÜ·. A±µ½¹º A±µ½F¯ªÌXebvÅ·. A±µ½¹ºðF¯Å«éuÍÅà ìèÉ¢àÌÅ·, ȺÈçbÌ«EðÁè·é½ßÉÁêÈû@ð gpµÈ¯êÎÈçÈ¢©çÅ·. A±µ½¹ºF¯uÍ[UÉÙÆ ñÇ©RÉb·±ÆðµÜ·, êûÅRs [^ªàeðÁèµÜ·. î{IÉ, »êÍRs [^Ì«æèÅ·. ©Rȹº ÀÛÉ©Rȹºª½Å é©Ìè`Í³Ü´Ü éæ¤Å·. î{IÈ iKÅÍ, »êÍ©Rȹ̶ŠÁÄJèÔ³êéàÌÅÍȢƢ ¤l¦Å é©àµêܹñ. ©RȹºÌ@\ðõ¦½ ASR VXeÍ "ums" Æ "ahs" ÈÇ, ¬³ê½¾tÈÇ̳ܴÜÈ©R̹ºÌÁ¥ â, ÷©Èû²à賦, µ¤±ÆªÂ\ŵå¤. ¹ºÆ/¯Ê ¢Â©Ì ASR VXeÍÁèÌ[Uð¯Ê·é@\ðÁĢܷ. ±Ì¶ÅÍÆâZL eB̽ßÌVXeÉ¢Ä͵¢Ü¹ñ. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 3.3. p@Æp Rs [^ÆlÔðî·édSÊɨ¢Ä, ASR ÌoÔª é©àµê ܹñ. »ÝͺLÉ°½AvP[VªêÊIÅ·. «æè «æèÍ, ¡úÅàêÊIÈ ASR VXeÌgp@Å·. ±êÍêÊÌ ¶Æ¯lÉãwL^]Êâ, @¥âdÌ«æèàÜÝÜ·. V Xe̸xðüã³¹é½ßÉ, ÁÊÈêbªgíêéêà èÜ·. ¹º½ßVXe Rs [^ÌR}hðÀs·é ASR VXe̱Æð, ¹º½ßVX eÆè`µÜ·. "Open Netscape" â "Start a new xterm" Ìæ¤É¹ ºÅ½ß·éÆ, bǨèÌR}hªÀs³êÜ·. db ¢Â©Ì PBX/Voice [VXeÍ, {^ð·©íèÉR}h ðb·±ÆÅdbð©¯çêÜ·. gÑ@í üÍèiªÀè³êÄ¢égÑ@íÅÍ, b·±ÆÍRÂ\Å·. ãÃ/nfBLbv ½Ìlª, ½^®ß½¹ (RSI), ØWXgtB[ÈÇÌæ¤Èg ÌIȧÀ̽ßÉ^CsOÉâèðø¦Ä¢Ü·. á¦Î®oÉâè Ì élÍ, è̺ðeLXgÉϦé½ßÉdbÉÚ±³ê½VXe ðgpÅ«éŵå¤. gÝÝAvP[V Vµ¢gÑdbÌÈ©ÉÍ "Call Home" Ìæ¤Èbððß·é C&C ¹ ºF¯ðõ¦½àÌà èÜ·. ±êÍ«Ì ASR Æ Linux Ìåv ÆÈé©àµêܹñ. ȺÍܾerÉbµ©¯çêÈ¢ÌÅµå ¤©? ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 4. n[hEFA 4.1. TEhJ[h ¹ºÍärIá¢ÑæðKvÆ·éÌÅ, öx©çi¿Ì 16 rbgT EhJ[hÈçg¦éŵå¤. J[lÅTEhðLøɵijµ¢h CoðCXg[µÈ¯êÎÈèܹñ. TEhJ[hÉ¢Ä̱ê ÈãÌîñÍ http://www.LinuxDoc.org/ É é "The Linux Sound HOWTO" ð ©Ä¾³¢. TEhJ[hÌi¿É¢Ä͸xÆmCYÌe¿É¢Ä, µÎµÎc_ªÜ«N±èÜ·. ÅàãYíÈ A/D (AiO©çfBW^) ÖÌÏ·@\ðÁ½TEhJ [hð©ßÜ·ª, µÎµÎfBW^Tv̾ijÍ}CNÌ«\ÉË ¶µ, üÍÌmCYÉÍ¢Á»¤å«Ë¶µÜ·. j^â, PCI Xbg, n[hfBXNÈÇ©çÌdCMIÈmCYÍÓ¤, Rs [^Ìt@ âÖqÌ«µÞ¹, Äz©ç·±¦émCYÉä×Ĭ³ÈàÌÅ·. ASR \tgEFApbP[WÉÍÁèÌTEhJ[hðKvÆ·éà̪ èÜ·. ÁèÌn[hEFAÖÌ˶ðð¯éÌÍÊíÇ¢±ÆÅ·, ÈºÈ ç«ÌIðð·ßĵܤ©çÅ·. àµKØÉ®ì·é½ßÉÍÁÊÈ n[hEFAªKvÆÈéæ¤ÈpbP[Wðl¦Ä¢éÌÈç, È½Í vÆRXgðl¶µÈ¯êÎÈçȢŵå¤. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 4.2. }CN }CNÌi¿Í ASR ðg¤ãÅdvÅ·. ½Ìêɨ¢Ä, ìã}CNÍ ±Ìgp@Éü«Ü¹ñ. üÍÌmCYðE¢ª¿ÉÈéÌÅ, ASR vO ª¤Ü®ìµÈ¢±ÆÉÈèÜ·. }CNð¸ÁÆ¿ÂïĢéÌÍåÏÈÌÅ, nh}CNàÅPÌIð ÅÍ èܹñ. üÍÌmCYÌÊð}¦Èªç, pÉÉbÒªÏíéêâ F¯uÉüÁÄb·±Æª ÜèÈ¢êÍ (wbhZbgðt¯é±Æà IðūȢƫ) ÅàÖÅ·. fRlCÌ éêÔæ¢IðÍwbhZbgÅ·. »êðg¦Î, ¢Âà È ½ÌûàÆÉ}CNðu¢½ÜÜÅà, üÍ̹ðŬÉ}¦é±ÆªÅ« Ü·. wbhZbgÍCAz̳¢àÌÆ éàÌ (m©XeI©) à èÜ·. XeIÌwbhzð©ßÜ·ª, »êÍÂlÌDÝÌâèÅ ·. $25 ©ç $100 ç¢Åf°çµ¢«\ðÁ½wbhZbg^}CNª¦ Ü·. http://www.headphones.com © http://www.speechcontrol.com ©çT µÄÝľ³¢. xÉ¢ÄÌZ©¢: }CNÌ{ [ðã°é±ÆðYêȢŠ¾³¢. ±êÍ XMixer © OSS Mixer Ìæ¤ÈvOðgÁÄsȤ± ƪūܷ, »µÄtB[hobNmCYðð¯éæ¤Égp·é±ÆÉ ÓµÄ¾³¢. ASR \tgEFAª©®²ßvOðÜñÅ¢êÎ, » êçðãèÉgÁľ³¢, »êçÍ»ÌÁèÌF¯VXeÉÅK»³ê Ģܷ. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 4.3. Rs [^/vZbT ASR AvP[VÍvZbT̬xÉ˶·é±Æª èÜ·. ± êÍ ASR ÅÍåÏÈÊÌfBW^tB^OÆMªN±è¤é© çÅ·. CPU ×Ì¢\tgEFAƯ¶, ¬¢ÙÇÇÈèÜ·. ܽ, ªå«¢öæÈèÜ·. ¢Â©Ì ASR Í 100MHz Æ 16MB Ì RAM Åà \Å·ª, ¬Å·é (å«È«â¡GÈF¯XL[, Tv [g) ÉÍ, ÅáÅà 400MHz Å 128MB Ì RAM ªÇ¢Åµå¤. KvÆ·é «\ÌÖWÅ, ÙÆñÇÌ\tgEFAÅÍŬÀÌKvðªLÚ³ê Ģܷ. åKÍÌF¯ðsȤÌÉ, NX^ (Beowulf â¼àÌà) ðp·é±Æ ÍsÈíêĢܹñ. isâJÌvWFNgð²¶mÈç¨mç¹ ¾³¢. scook@gear21.com <mailto:scook@gear21.com> ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5. ¹ºF¯\tgEFA 5.1. t[\tgEFA ±±Å°ét[\tgEFA̽Í, ±±©ç_E[hÅ«Ü·: http://sunsite.uio.no/pub/Linux/sound/apps/speech/ ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.1.1. XVoice XVoice ͳܴÜÈ XWindow AvP[VÅgpū鹺F¯Ì\t gEFAÅ, «æèâA±µ½¹ºF¯ªÂ\Å·. [Uª}Nðè` ·é±ÆàÅ«, m©È¢Ì éÇ¢vOÅ·. êxÝè·êÎ, [ ªÈ¸xÅ®ìµÜ·. XVoice ðg¤½ßÉÍ IBM Ì ViaVoice for Linux (¤pÌßð©Ä¾³ ¢) ðüèµÄCXg[·éKvª èÜ·. ܽ ViaVoice ð³µ® ì³¹é½ßÉÝèªKvÅ·. Á¦Ä Lesstif/Motif (libXm) àKvÅ·. ±ÌvOÍ X Window ÆâèÆè·éÌÅ, X \[XðpÅ«éæ ¤ÉµÄ¨©È¯êÎÈçÈ¢±ÆÉÓ·é±ÆàdvÅ·, ±Ì½ß, l bg[NÉpªÁ½}Vâ}`[UÌ}VÅgp·éÆ«Í, Cð t¯Ä¾³¢. ±Ì\tgEFAÍåÉ[Uü¯Å·. RPM ðüèÅ«Ü·. HomePage: http://www.compapp.dcu.ie/~tdoris/Xvoice/ http:// www.zachary.com/creemer/xvoice.html Project: http://xvoice.sourceforge.net Community: http://www.onelist.com/community/xvoice ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.1.2. CVoiceControl/kVoiceControl CVoiceControl (Console Voice Control ̪) ͳX KVoiceControl(KDE Voice Control) ŵ½. ±ÌvOÍ[UªR}hðb·±ÆÅ Linux ÌR}hðÀsÅ«é, î{IȹºF¯VXeÅ·. CVoiceControl ª KVoiceControl Éu«ãíèܵ½. ±Ì\tgEFAÉÍ}CNxðÝè·é[eBeB, Vµ¢R} hÆbðÇÁ·é½ßÌêbfGfB^, ¹ºF¯VXeªÜÜêÄ ¢Ü·. CVoiceControl Í ASR ðnßæ¤Æ·éo±LxÈ[UÉÆÁÄ, f°çµ ¢o_ÆÈèÜ·. K¸µà[UthÅ éÆ;¦Ü¹ñª, ³ µwK³¹êÎ, ÆÄàð§¿Ü·. ZbgAbvðs¤ÉÍhL gðÇÇñž³¢. ±Ì\tgEFAÍåÉ[Uü¯Å·. Homepage: http://www.kiecza.de/daniel/linux/index.html Documents: http://www.kiecza.de/daniel/linux/cvoicecontrol/index.html ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.1.3. Open Mind Speech 1999 Nã¼ÉnÜÁ½ Open Mind Speech ͽx©¼OðϦܵ½ (©ÂÄ Í VoiceControl, »ÌãÍ SpeechInput Å, »ê©ç FreeSpeech Å·). Ü ½¡ÅÍ, I[v\[XvWFNgÌ "Open Mind Initiative" ÌêÅ ·. ¡ÌÍ®SÉ@\·éí¯ÅÍÈ, åÉJÒü¯Å·. ±Ì\tgEFAÍåÉJÒÉü¯½àÌÅ·. Homepage: http://freespeech.sourceforge.net ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.1.4. GVoice GVoice Í Gtk/GNOME AvP[Vð§ä·é½ßÉ IBM Ì (t[Ì) ViaVoice SDK ðgpµ½¹º ASR CuÅ, ú», F¯GW, « ì, plÌRg[ðs¤½ßÌCuªÜÜêĢܷ. J ÍêNÈãâصĢܷ. ±Ì\tgEFAÍåÉJÒÉü¯½àÌÅ·. Homepage: http://www.cse.ogi.edu/~omega/gnome/gvoice/ ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.1.5. ISIP Mississippi State University Ì Institute for Signal and Information Processing ͻ̹ºF¯GWðöJµÜµ½. ±Ìc[LbgÍt gGhÆfR[_[, »µÄPûW [ðÜñŢܷ. ±êÍ@\ IÈc[LbgÅ·. ±Ì\tgEFAÍåÉJÒÉü¯½àÌÅ·. ±Ìc[Lbg (Æ ISIP É¢ÄÌîñ) ͱ±ÅüèÅ«Ü·: http:// www.isip.msstate.edu/project/speech/ ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.1.6. CMU Sphinx Sphinx ÍàÆàÆ CMU Ånßçê, ÅßI[v\[XƵÄöJ³êܵ ½. ±êͽÌc[ÆîñðÜñ¾, ©Èèå«ÈvOÅ·. ±ê ܾ͢É"J"Å·ª, wK̽ßÌ\tgEFAÆF¯u, ¹¿f , ¾êf, ì¬Ì¶ðÜñŢܷ. ±Ì\tgEFAÍåÉJÒÉü¯½àÌÅ·. Homepage: http://www.speech.cs.cmu.edu/sphinx/Sphinx.html Source: http://download.sourceforge.net/cmusphinx/sphinx2-0.1a.tar.gz ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.1.7. Ears Ears ÌJÍ®SÅÍ èܹñª, ASR ðnß½¢ÆvÁÄ¢évO} ÉÍÇ¢«Á©¯ÉÈéŵå¤. ±Ì\tgEFAÍåÉJÒÉü¯½àÌÅ·. FTP site: ftp://svr-ftp.eng.cam.ac.uk/comp.speech/recognition/ ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.1.8. NICO ANN Toolkit NICO Artificial Neural Network toolkit͹ºF¯AvP[VÉÅK »³ê½tLVuobNvpQ[Vj [lbg[Nc[ LbgÅ·. ±Ì\tgEFAÍåÉJÒÉü¯½àÌÅ·. homepage: http://www.speech.kth.se/NICO/index.html ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.1.9. Myers' Hidden Markov Model Software Richard Myers ̱Ì\tgEFAÍ C++ ÅLq³ê½ HMM ASYÅ ·. ±êÍ L. Rabiner Ì{Å é "Fundamentals of Speech Recognition" ÉLq³ê½ HMM ̽ßÌáÆwKc[ðñµÜ·. ±Ì\tgEFAÍåÉJÒÉü¯½àÌÅ·. îñͱ±ÅüèÅ«Ü·: http://www.itl.atr.co.jp/comp.speech/Section6 /Recognition/myers.hmm.html ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.1.10. Jialong He's Speech Recognition Research Tool àÆàÆ Linux ü¯É©ê½àÌÅÍ èܹñª, ±Ì¤c[Í Linux ÅRpCÅ«Ü·. ÙÈé3ÂÌ^CvÌF¯uðõ¦Ä¢Ü·: DTW, Dynamic Hidden Markov Model, Continuous Density Hidden Markov Model Å·. ±êͤÆJpÌàÌÅ, ®SÈ ASR VXeÅÍ èܹ ñ. ±Ìc[Lbg͢©ÌÖÈc[ðÁĢܷ. ±Ì\tgEFAÍåÉJÒÉü¯½àÌÅ·. ³çɽÌîñͱ±ÅüèÅ«Ü·: http://www.itl.atr.co.jp/ comp.speech/Section6/Recognition/jialong.html ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.1.11. ܾ¼Éà èÜ·©? ãLÈOÌàÌ𲶶ÈçÜŨmç¹¾³¢: scook@gear21.com <mailto:scook@gear21.com>. àµæ뵩Á½ç, Ð½\tgEFAÌ Rs[ðüèÅ«éê೦ľ³¢. ³çÉ´zàÁĸ¯éÆK¢ Å·. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.2. ¤p\tgEFA 5.2.1. IBM ViaVoice SDK Ì¢ÍǤÈé©í©èܹñª, IBM Í ViaVoice V[YÅ Linux ðT|[g·éÆ¢¤ñ©ðµÄ¢Ü·, (JÒÆÌCZX_ñÍ»_ ÅÍö®ÉÍsíêĢܹñ, वæÉÈéŵå¤. ) ¤pÌ (t[ÅÈ¢) »iÅ é, IBM ViaVoice Dictation for Linux (http://www-4.ibm.com/software/speech/linux/dictation.html ©çüèÅ« Ü·) Ì«\ÍÆÄàÇ¢ÌÅ·ª, î{IÈ ASR VXe (64M RAM Æ 233MHz Pentium) ÉärµÄ³çÉå«ÈVXeðKvƵܷ. $59.95US Å Andrea NC-8 }CNàt®µÄ¢Ü·. }`[UÅgp·é±ÆàÂ\ Å· (µ©µ, Í}`[UÅÀ±µÄ¢È¢ÌÅ, ¾ê©À±µ½lª ¢êĮ̂µèð@¢Ä¾³¢). ±ÌpbP[WÍÌàÌðÜÝÜ·: ¶ (PDF), wKc[, «æèVXe, »ê©çCXg[XNv g. 2.2nJ[lðx[Xɵ½¼Ì Linux fBXgr [VÌT| [gàÅVÌ[XÅͳêĢܷ. ±Ì ASR SDK Í©RÉüèÅ«, IBM Ì SMAPI, ¶@ API, ¶, ÆlXÈT vvOðÜñŢܷ. ViaVoice Run Time Kit Í«æè@\Ì ½ßÌ ASR GWÆf[^t@C, [U[eBeBðñµÜ·. ±Ì ViaVoice Command & Control Run Time Kit ͹º½ßVXe̽ßÌ ASR GWÆf[^t@CÆ[U[eBeBðÜñŢܷ. ±Ì SDK Æ Kit ÉÍ 128MB Ì RAM Æ Linux 2.2 ÈãªKvÅ·. SDK Æ Kit ͱ±Å©RÉüèÅ«Ü·: http://www-4.ibm.com/software/ speech/dev/sdk_linux.html ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.2.2. Vocalis Speechware Vocalis Æ Vocalis Speechware É¢Ä̳çÈéîñÍ: http:// www.vocalisspeechware.com Æ http://www.vocalis.com. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.2.3. Babel Technologies Babel Technologies Í Babear ÆÄÎêé Linux SDK ðñµÄ¢Ü·. ± êÍ Hybrid Markov Model Æ Artificial Neural Network eNmWÉîà ¢½bÒÉ˶µÈ¢VXeÅ·. eLXg¹ºÏ·âbÒÆ, ¹fðÍ ÉÖ·é³Ü´ÜÈ»iª èÜ·. ¼ÌîñÉ¢ÄÍ: http:// www.babeltech.com. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.2.4. SpeechWorks ÞçÌEFuTCgÅÍ Linux É¢ÄÁɾyµÄ¢Ü¹ñª, ÞçÌ "OpenSpeech Recognizer" ÍI[vX^_[hÅ é VoiceXML ðgpµ Ģܷ. ¼ÌîñÉ¢ÄÍ: http://www.speechworks.com. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.2.5. Nuance Nuance ͳܴÜÈ *nix vbgtH[p̹ºF¯/©R¾êÌ»i (»ÝÍ Nuance 8.0) ðñµÄ¢Ü·. ñíÉå«Èêb𵤱ƪÂ\ ÅXP[reBÆáQe̽ßÉÁL̪UA[LeN`ðgpµÄ ¢Ü·. ¼Ìîñͱ±©çüèÅ«Ü·: http://www.nuance.com. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.2.6. Abbot/AbbotDemo Abbot ÍñíÉå«Èêb, bÒÉ˶µÈ¢ ASR VXeÅ·. »êÍàÆ àÆ, Cambridge University Ì Connectionist Speech GroupÉæÁÄJ³ ê, ¢ÜÍ, SoftSound (¤p)ÉÚÁĢܷ. íµ¢îñÍ: http:// www.softsound.com AbbotDemo Í Abbot ÌfpbP[WÅ·. ±ÌfVXeÍñ 5000 êÌ êbð¿, connectionist/HMM ÌA±µ½¹ºASYðpµÄ¢Ü ·. ±êÍ\[XR[hÌ®µÈ¢fvOÅ·. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.2.7. Entropic Entropic ÌüÓÌL\ÈlÍ Micro$oft É¢æçêĵܢܵ½. . . »iÆT|[gT[rXÍSÄÁ¦ÄµÜ¢Üµ½. HTK Æ ESPS/waves+ ÌT |[gÍÅ¿Øçêĵܢ, ÞçÌ¢Í M$ É©©ÁĢܷ. âEF uTCg http://www.entropic.com ɳçÉîñª èÜ·. K.K. Chin ªÉ HTK ̳XÌJÒ (Cambridge Ì Speech Vision and Robotic Group) ªÜ¾»êÉηéT|[gðµÄ¢éÆîñðêܵ½. http://htk.eng.cam.ac.ukÅÍt[Èo[WàüèÅ«Ü·. Microsoft ª»sÌ HTK ÌR[hÌì ðLµÄ¢é±ÆÉÍӵľ³¢. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 5.2.8. ¼Ì¤p»i æè½Ì¤pÌ ASR »iª (L&H ðÜßÄ) ߢ«üèÅ«éæ¤ÉÈé Æ¢¤\ª èÜ·. Í Comdex 2000 (Vegas) Å L&H Ìã\ 2,3 lÆbð µÜµ½ª, Nà Linux [XÉ¢Ä, ܽ Linux ü¯ÉÇÌ»iÌ [Xðv浽̩É¢ij¦îñðêܹñŵ½. ൱êÈãÌ îñðÁÄ¢êÎ, Ú×ð scook@gear21.com <mailto:scook@gear21.com> ÉÁľ³¢. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 6. ¹ºF¯Ìठ6.1. ÇÌæ¤ÉF¯µÄ¢é© F¯VXeÍ 2 ÂÌåȪɪðÅ«Ü·. p^[F¯VXeÍK ð»f·é½ßÉp^[ðùmÌàÌâwKµ½p^[ÆärµÜ·. Acoustic Phonetic VXe͹ºÌÁ¥ (ê¹ÈÇ̹ÈÇ̹º) ðär ·é½ßÉlÌÉÖ·ém¯ (ºÌ¶¬, Æ®o) ðpµÜ·. ÙÆñÇÌ »ãIÈVXeͱÌæ¤Èp^[F¯Av[`Éd_ðu¢Ä¢Ü·, ȺÈç. »êÍ»ÝÌRs [^pZpƤÜÑ«, ¢¸xð ¾â·¢©çÅ·. ÙÆñÇÌF¯uÍȺÌæ¤ÈiKɪðÅ«Ü·: 1. I[fBIÌL^ÆbÌo 2. vtB^O (vGt@TCY, ³K», ofBOÈÇ) 3. t[~OÆEBhEBO (f[^ðgpÅ«é`®Éªð·é) 4. tB^O (XÉ»ê¼êÌwindow/frame/freq.bandðtB^ O) 5. ärÆK (bÌF¯) 6. ®ì (F¯³ê½p^[ÉÖAµ½@\ðÀs) »ê¼êÌiKÍPÉ©¦Ü·ª, êÂêÂͽÌÙÈé (»µÄÆ«Ç «Í®SÉtÌ) ZpðpµÄ¢Ü·. (1) I[fBI/ºÌ^¹: ¢ë¢ëÈû@ª èÜ·. nßÍüÍÌI[f BIÌx (¢Â©ÌêÅ͹¿ÌGlM[) ð^¹³êÄ¢éT vÆär·é±ÆÅ·. I_Ì»ÊÍ, bÒªÄzâ½ß§, Ì«, G R[ÈÇÌ "artifacts" ðcµª¿ÈÌųçÉ¢ïÅ·. (2) vtB^O: F¯VXe̼Ì@\É˶µÄ, ¢ë¢ëÈû @ÅsÈíêÜ·. ÅàêÊIÈû@Í, TvÌõ̽ßÉêAÌI[ fBItB^[ðgp·é "Bank-of-Filters" @Æ, ·Ù(ë·)ÌvZ̽ ßÉ\ª@\ðgp·é Linear Predictive Coding @Å·. ÙÈé`®ÌX yNgðÍàp³êÜ·. (3) t[~O/EChEBOÍTvf[^ðÁèÌ嫳ɪ· é±ÆÅ·. ±ê͵εΠstep2 Æ step4 ÖiÝÜ·. ±ÌiKͪÍÌ ½ßÉTv«E (£ÌJ`Æ¢¤¹ÈÇð±Æ) ðpÓ·é±Æà ÜñŢܷ. (4) ÇÁÌtB^OÍ¢Âà¶Ý·éí¯ÅÍ èܹñ. ±êÍär ÆKÌOÅÌ»ê¼êÌEBhEÉηéÅãÌõÅ·. µÎµÎ±ê ÍÔÌzuƳK»©ç\¬³êÜ·. (5)ÌärÆKÉ¢ÄÍÂ\ÈZpªåÊÉ èÜ·. ÙÆñÇÍ»ÝÌE BhEÆùmÌTvÌärðKvƵܷ.Hidden Markov Models (HMM), ügðÍ, ·ÙðÍ, ü^ãÌZp/ß¹, XyNgcðp· éû@âÔcÈ@à èÜ·. ±êçÌ·×ÄÌû@ÍêvÌm¦Æ¸xð ¾é½ßÉgp³êĢܷ. (6) ®ìÍJÒª]ñ¾±Æ¾¯Å·. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 6.2. fBW^I[fBIÌîb I[fBIÍ{¿IÉAiOÈ»ÛÅ·. fBW^TvÅ^¹·é± ÆÍ, }CN©çÌAiOMðTEhJ[hãÌ A/D Ro[^ÅfB W^MÉÏ··é±ÆÅ·. }CNª®ìµÄ¢êÎ, ¹gÍ}CNÌ Ì¥ÎÌvfðU®³¹, TEhJ[hÖÌd¬ (Xs[JªtÉ®ìµÄ ¢éÆl¦Ä¾³¢) 𶳹ܷ. î{IÉ»ÌA/DRo[^ÍÁèÌ ÔuÅÌd³ÌlðL^µÜ·. ±ÌßöÌÉ2ÂÌdvÈvfª èÜ·. 1ÂßÍ "sample rate", é¢ ÍÇÌæ¤ÈpxÅd³ðL^·éÌ©Æ¢¤àÌ. 2ÂßÍ "bits per second", ÇÌæ¤È¸xÅlªL^³êé©Æ¢¤àÌÅ·. 3ÂßÌvfÍ `lÌ (m©XeI©), µ©µ, ½Ì ASR AvP[V ÅÍmÅ\ªÅ·. ½ÌAvP[VÅͱêçÌp[^É\ß Ýè³ê½lðgpµÄ, [UͶɩêĢȢ©¬èÏX·é׫ ÅÍ èܹñ. JÒÍÙÈélÅ»ÌASYÆͽª¤Ì©ðÀ ±·é±ÆÅè·é׫ŷ. »êÅÍ, ASR É¢ÄÍÇÌæ¤ÈTv[gªÇ¢Ìŵ天? ¹ ºÍärIá¢Ñæ (ÙÆñÇ 100Hz ©ç 8kHz) Å·©ç, 8000 samples/ sec (8kHz) ÍÙÆñÇÌî{IÈ ASR ÉεÄÍ\ªÅ·. µ©µ, ³çÉ ³mÈügÌîñð¾çêéÌÅ 16000 samples/sec(16kHz) ðDÞlà ¢Ü·. àµ\ͪ êÎ 16kHz ðg¤×«Å·. ÙÆñÇÌ ASR Av P[VÅÍ 22kHz ÈãÌTvO[gͳÊÅ·. »µÄÇÌæ¤Èlª "bits per sample" (1TvèÌrbg) É¢ ÄÇ¢Ìŵ天? 8 bits per sample Í 0 ©ç 255 ÌÔÅlðL^µÜ ·, ±êÍ}CN¬ªÌ嫳ª 256ÂÌÌ1ÂÅ éÆ¢¤±ÆðÓ¡µÜ ·. 16 bits per sampleͬªÌ嫳ð 65536 ÂɪµÜ·. Tv [gà¯lÅ·. är̽ßÉ, ¹yp Compact Disc Í 44kHz Å 16 bits per sampleÅGR[h³êĢܷ. gp³êéGR[fBOtH[}bgÍü^Èt é¢Íȵ® Ìæ¤ÉPÅ é׫ŷ. U-Law/A-Law ASYܽͼ̳k@ ðgp·é±ÆÍÊ¿lª èܹñ, ȺÈç»êÍvZ\ÍÌRXgª ©©è, RXgÉ©Á½\Íð\ªÉ¾é±ÆªÅ«È¢©çÅ· ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 7. oŨ ±ÌXgÉÚÁĢȢàÌÅ, ±±ÉÁ¦½Ù¤ª¢¢Æv¤oŨª êÎ, scook@gear21.com <mailto:scook@gear21.com>ÖîñðÁľ³ ¢. ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 7.1. Ð E "Fundamentals of Speech Recognition". L. Rabiner & B. Juang. 1993. ISBN: 0130151572. E "How to Build a Speech Recognition Application". B. Balentine, D. Morgan, and W. Meisel. 1999. ISBN: 0967127815. E "Speech Recognition : Theory and C++ Implementation". C. Becchetti and L.P. Ricotti. 1999. ISBN: 0471977306. E "Applied Speech Technology". A. Syrdal, R. Bennett, S. Greenspan. 1994. ISBN: 0849394562. E "Speech Recognition : The Complete Practical Reference Guide". P. Foster, T. Schalk. 1993. ISBN: 0936648392. E "Speech and Language Processing: An Introduction to Natural Language Processing, Computational Linguistics and Speech Recognition". D. Jurafsky, J. Martin. 2000. ISBN: 0130950696. E "Discrete-Time Processing of Speech Signals (IEEE Press Classic Reissue)". J. Deller, J. Hansen, J. Proakis. 1999. ISBN: 0780353862. E "Statistical Methods for Speech Recognition (Language, Speech, and Communication)". F. Jelinek. 1999. ISBN: 0262100665. E "Digital Processing of Speech Signals" L. Rabiner, R. Schafer. 1978. ISBN: 0132136031 E "Foundations of Statistical Natural Language Processing". C. Manning, H. Schutze. 1999. ISBN: 0262133601. ½ÌICÅÇßéãLª éÌÅ, Institut Fur Phoneti ð`FbN µ½Ù¤ª¢¢Åµå¤: http://www.informatik.uni-frankfurt.de/~ifb/ bib_engl.html ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 7.2. C^[lbg news:comp.speech Rs [^ƹºÉÖ·éj [XO[vÅ·. US: http://www.speech.cs.cmu.edu/comp.speech/ UK: http://svr-www.eng.cam.ac.uk/comp.speech/ Aus: http://www.speech.su.oz.au/comp.speech/ news:comp.speech.users ¹ºÉÖ·é\tgEFAÌ[U̽ßÌj [XO[vÅ·. http://www.speechtechnology.com/users/comp.speech.users.html news:comp.speech.research ¹ºÉÖWµ½\tgEFAÆn[hEFA̽ßÌj [XO[vÅ ·. news:comp.dsp fBW^M̽ßÌj [XO[vÅ·. news:alt.sci.physics.acoustics ¹Ì¨w̽ßÌj [XO[vÅ·. DDLinux Email List Linux ̹ºF¯Ì[OXgÅ·. Homepage: http://leb.net/ddlinux/ Archives: http://leb.net/pipermail/ddlinux/ Linux Software Repository for speech applications http://sunsite.uio.no/pub/linux/sound/apps/speech/ Russ Wilcox's List of Speech Recognition Links (excellent) http://www.tiac.net/users/rwilcox/speech.html Online Bibliography Online Bibliography of Phonetics and Speech Technology Publications. http://www.informatik.uni-frankfurt.de/~ifb/ bib_engl.html MIT's Spoken Language Systems Homepage http://www.sls.lcs.mit.edu/sls/ Oregon Graduate Institute Oregon Graduate Institute Ì Spoken Language Understanding ÌZ^ [Å·. JÒƤÒÉÆÁÄf°çµ¢êÅ·. http:// cslu.cse.ogi.edu/ IBM's ViaVoice Linux SDK http://www-4.ibm.com/software/speech/dev/sdk_linux.html Mississippi State Signal and Information ProcessingÉ¢ÄÌ~VVbsB§åwJÒ Éü¯½åÊÌîñª éz[y[WÅ·. http:// www.isip.msstate.edu/projects/speech/ Speech Technology ASR \tgEFAÆANZTÅ·. http://www.speechtechnology.com Speech Control ¹º§äÌRs [^VXe. ASRpÌ}CN, wbhZbg, C X»iÅ·. http://www.speechcontrol.com Microphones.com ASR pÌ}CNÆANZTÅ·. http://www.microphones.com 21st Century Eloquence "Speech Recognition Specialists." http://voicerecognition.com Computing Out Loud åÉÍ Windows [UÉü¯½àÌÅ·ª, Ç¢îñª èÜ·. http:/ /www.out-loud.com Say I Can.com "The Speech Recognition Information Source." http://www.sayican.com ªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªªª 8. ú{êóÉ墀 ú{êóÍ Linux Japanese FAQ Project ªs¢Üµ½. |óÉÖ·é²Ó© Í JF vWFNg <JF@linux.or.jp> ¶ÉAµÄ¾³¢. 1.2j |ó: <htakashi@yabumi.com> Z³: JçG <jeanne@mbox.kyoto-inet.or.jp> ì{_ê <hng@ps.ksky.ne.jp>