The usage of multiple Hidden Markov Models (HMMs) to prepare a Czech acoustic unit inventory and speech synthesis based on this inventory are presented in this paper ( Triphone HMMs are trained on the basis of the speech corpus spoken by a single speaker.
)LJXUH LQYHQWRU\ $Q LOOXVWUDWLRQ RI
VSHHFK FRQVWUXFWLRQ

FRUSXV FRQVWUXFWLRQ DQ DXWRPDWLF DFRXVWLF XQLW
VHJPHQWDWLRQ LQFOXGHV DQG +00 VHJPHQW WUDLQLQJ GDWDEDVHDQG
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
RI FOXVWHUHG LQ RXU DSSUR[LPDWHO\
VWDWHV WUDLQLQJ ZDV GDWD
DERXW $IWHU FOXVWHULQJ WULSKRQHV
,Q WKH
1R VLQFH YHULILFDWLRQ
ODUJH FRUSXV LW LV YHU\ RI ODERULRXV WKH VHJPHQWDWLRQ WR PDQXDOO\ UHVXOWV VHJPHQW ZDV VXFK GRQHD
63((&+ 6(*0(17 '$7$%$6(&216758&7,21
6SHHFK DUH VHJPHQWV FRUUHVSRQGLQJ WR HDFK FOXVWHUHG VWDWHVHYHUDO VWRUHG
DSSURDFK LV WKDW RQO\ RQH VHJPHQW LV VHOHFWHG WR UHSUHVHQWZD\V LQ WKH KRZ VSHHFK WR FRQVWUXFW VHJPHQW GDWDEDVH WKH GDWDEDVH 7KHUH 2QHDUHHDFK JRRG RU DYHUDJH HQRXJK WR EH XVHG DQ\WLPH WKH GHVLUHGFOXVWHUHG VWDWH 7KH VHOHFWHG VHJPHQW VKRXOG EHVRXQG L H DFRXVWLF XQLW LV WR V\QWKHVL]H
$QRWKHU ZD\ KRZ WR FRQVWUXFW WKH GDWDEDVH LV WKDW PRUHVHJPHQWV VWRUHG GXULQJ '\QDPLF RU DOO SURJUDPPLQJ RFFXUUHQFHV SHU FRXOG FOXVWHUHG VWDWH DUHWKH VRXQG LV VHOHFWHG 7KH SURJUDPPLQJ LV XVXDOO\ EDVHG RQSKRQHWLF V\QWKHVLV DQG HQVXULQJ SURVRGLF WKDW FRQWH[W VHJPHQW EH
RI WR EH V\QWKHVL]HGZKLFK WKHQ DSSOLHGEHVW ILW
D FRVW IXQFWLRQ WKDW HYDOXDWHV SURVRGLF IHDWXUHV DQGVSHFWUDO VHJPHQWV > @ +HUH DJDLQ FRPSURPLVHV EHWZHHQ PHPRU\GLVFRQWLQXLW\ RI WKH DGMDFHQW WR EH V\QWKHVL]HGUHVRXUFHV UXQWLPH DQG VSHHFK TXDOLW\ FDQ WDNHQ LQWR DFFRXQW V\QWKHVLV WKH UXQWLPH VSHHG EH PDGH VKRXOG ,Q DOVR FDVH EHRI,Q DSSURDFK 7KH PDLQ FULWHULD IRU VHOHFWLQJ D UHSUHVHQWDWLYHRXU FXUUHQW LPSOHPHQWDWLRQ ZH XVH WKH IRUPHUVHJPHQW IURP VHJPHQWHG VSHHFK FRUSXV LV D +00 VFRUHLPSRVHG VRPH ZD\ H[WUHPH VHJPHQW H J VHJPHQW ZLWK WRR ELJ RUE\ 9LWHUEL VHDUFK 7R DYRLG VHOHFWLQJ DQ LQVPDOO DQG GXUDWLRQ VWDWLVWLFV ZHUH FDOFXODWHG HQHUJ\ RU WRR ORQJ RU VKRUW VHJPHQW DPSOLWXGHXQLW IRU HDFK DFRXVWLFSDUWLFXODU L H FOXVWHUHG ZKRVH FOXVWHUHG VWDWH VWDWH LQ IURP WKH DOO WUDLQLQJ WKH RFFXUUHQFHV RI DXQLW EH PHDQ DPSOLWXGHV DUH QRW DQG WDNHQ GXUDWLRQV LQWR DFFRXQW DUH IDU GDWD 3LWFK DZD\ 6HJPHQWVVKRXOG IURP DOVRWKHQRW XVH DQ\ SLWFK WUDFNLQJ DOJRULWKP WDNHQ LQWR DFFRXQW EXW FXUUHQW LPSOHPHQWDWLRQ GRHV 63((&+ 6<17+(6,67KHUH FRQFDWHQDWLRQ EDVHG VSHHFK V\QWKHVLV DUH VHYHUDO WHFKQLTXHV XVHG LQ FDOOHG GLUHFW V\QWKHVLV RU V\QWKHVLV EDVHG RQ 7KH WKH VLPSOHVW FRQWH[W RIFRGLQJ LV EDVHG RQ VLPSO\ FRQFDWHQDWLQJ ZDYHIRUPV RIZDYHIRUPRQHVSHHFK DQDO\VLV VHJPHQWV VSHHFK QRU SURVRGLF IURP VSHHFK PRGLILFDWLRQ GDWDEDVH 1HLWKHU VSHHFKLQWHOOLJLEO\ VHJPHQWV VWURQJO\ QR VSHHFK 7KH SDUDPHWHUL]DWLRQ UHVXOWLQJ LV VSHHFK SHUIRUPHG RQLV GRQH VRXQGVIHDWXUHV XQQDWXUDOO\ VHJPHQWV ZLWK GLIIHUHQW SURVRGLFEXWPRGLILFDWLRQ DUH VLPSO\ MRLQHG ZLWK QR SURVRGLF7KH SUHGLFWLYH FODVVLFDO /3 DSSURDFK V\QWKHVLV WR ,W VSHHFK LV EDVHG V\QWKHVLV RQ DXWRUHJUHVVLYH

LV OLQHDU
Adaptation of pitch and spectrum for HMM-based speech synthesis using MLLR...based TTS system in which each speech synthesis unit is modeled by HMM [...
Corpus-based approaches to speech synthesis have been advocated to overcome the limitations of concatenative synthesis from a xed acoustic unit inventory. ...
USING HMM-BASED SPEECH RECOGNITION/SYNTHESIS ...cient vectors for each speech unit is modeled by...
Speech Technologies Acoustic Modeling of Speaking ...synthetic speech using HMM-based speech synthesis....speech parameters of a speech unit such as the ...
automatically, using HMM-based approach [12, 13]....: Join Cost for Unit Selection Speech Synthesis...Kavai, H., Tsuzaki, M.: Acoustic Measures vs....
Speech Synthesis Using H... 5页 免费 An HMM... HMM-BASED SMOOTHING FOR CONCATENATIVE SPEECH ...acoustic quality of the Whistler Text-to-Speech ...
An HMM-based speech synt... 暂无评价 4页 免费 HMM-Based Speech Synthes... 185页 免费 Speech Synthesis Using H... 5页 免费 HMM-BASED SMOOTHING ...
RULE-BASED VISUAL SPEECH SYNTHESIS Jonas Beskow ...speech, especially under degraded acoustic conditions...modalities to be written using the same notation....
called HMM-based generation synthesis (or simply...All of these techniques depend on a acoustic ...Ho?man, “Speech synthesis using stochastic Markov...
HMM-Based Speech Synthes... 185页 免费 Speech Synthesis Using H... 5...(for which the acoustic evaluation is based on an Euclidean distance measure...
热门文档