 |
Description: Speech synthesis,prosody control, application of the
artificial neural networks
Jana Tuckova - Curriculum Vitae
Education:
1974 Graduated at the Czech Technical University in Prague (CTU)
1980 CSc thesis (PhD equivalent) at CTU Prague on
"Sensitivities of the Ladder Circuits" (in Czech)
1993 Postdoctoral course Biological and Artificial Neural
Networks at the Department of Informatics of the Swiss Federal
Institute of Technology in Lausanne, on "Vector Quantizer using
Kohonen Neural Network for the Speech Signal" (in French) (responsible
Prof. Martin Hasler)
1997 Habilitation thesis at CTU Prague on "Artificial Neural
Network and the Recognition of Isolated Words" (in Czech)
1998 Short-Term Scientific Mission, COST 258 , at TIK Computer
Engineering Laboratory, ETH, Zurich
Employment:
since 1998 Associate Professor in the Department of Circuit Theory
at CTU Prague
since 1995 Research in the Institute of Radio Engineering and
Electronics Academy of Sciences of the Czech Republic,Prague,
Department of Digital Signal Processing
1980-1997 Assistant Professor in the Department of Circuit Theory at
CTU Prague
1976-1980 Assistant in the Department of Circuit Theory at CTU
Prague
1974-1976 Doctoral study at the Institute for the communication
technology in Prague
Teaching at CTU Prague:
Seminar, laboratory and CAD Exercises on Circuit theory, on Linear
circuits and systems, on Electrical Filters
Lectures and CAD Exercises on Special structures of digital systems
Teaching at the Technical University of Brno:
Doctoral study courses Application of artificial neural
networks for speech processing.
Teaching at the Technical University of Zilina, Slovakia:
Lectures on Neural networks
Membership:
Member of International Neural Networks Society (INNS)
U.R.S.I. Correspondent (1999)
Member of International Program Committee and TechnicalCommittee of
IASTED International Conference SIP'99, SIP'2000, SIP'2001
Member of International Program Committee of IASTED International
Conference on Neural Networks 2000
A member of the IASTED Technical Committee on "Neural Networks" for
2001-2004"
Research topic
I work in the domain of artificial neural networks, data mining,
and speech synthesis. In 1992 I completed the six month research
fellowship on the CIRC-EPFL in Lausanne with Prof. Martin Hasler. At
time I started my research in speech processing. In 1993 I
obtained the Attestation from the Computer Science Postgraduate
Course of Biological and Artificial Neural Networks on EPFL. Now, I
concentrate on prosody modelling for Text-to-Speech Synthesis
Systems in Czech language especially. I applied the artificial
neural network approach for prosody parameters control.
75 contributions in the journals and proceedings were published in
1975 - 2001 ( in 1975-1993 the topics were from the theory of the
linear circuits, active filters especially, from 1993 the
publications are from the theory and applications of the artificial
neural networks and speech processing).
Main publications from 1996:
Proceedings of the International Conferences:
[ 1] Vich,R.,Tuckova,J.: Neural Network Application in Czech
Text-to-Speech Synthesis.
In:: Proc. of the IASTED Internat. Conf. Signal and Image Processing
and Applications
SIPA 96, Annecy, France, June 12-14, 1996, pp.242-244. ISBN"
0-88986-203-6.
[ 2] Tuckova,J., Vich,R.: Fundamental Frequency Control in Czech
Text-to-Speech Synthesis by a Multilayer Neural Network. In:
Abstracts of the XXVth General Assembly URSI Lille, France, August
28 - September 5, 1996, pp.135.
[ 3] Horak,P.,Tuckova,J.,Vich,R.: New Prosody Modelling System for
Czech Text-to-Speech. In: D.Mehnert (ed): Studientexte zur
Sprachkommunikation No. 13-Elektronische Sprachsignalverarbeitung,
Berlin, Nov.25.-27, 1996, pp.181-186.
[ 6] Tuckova,J.,Vich,R.: Fundamental Frequency Modelling by Neural
Nets in Czech Text-to-Speech Synthesis. Proc. of the IASTED Int.Conf.
" Signal and Image Processing-SIP'97, New Orleans, Louisiana, USA,
December 1997, pp.85-88. ISBN: 0-88986-242-7, ISSN: 1482-7921.
[ 7] Tuckova, J.: Report on Short-Term Scientific Mission. Management
Committee of COST 258, April 8-10,1999, Morges, Switzerland.
[ 8] Sebesta,V., Tuckova, J.: Selection of Important Input Parameters
for a Text-to-Speech Synthesis by Neural Networks. 1999 International
Joint Conference on Neural Networks IJCNN'99, Washington,DC., July
10-16, 1999. IEEE Catalog Number: 99CH36339C, ISBN: 0-7803-5532-6.
Proceedings on CD ROM.
[ 9] Tuckova,J., Sebesta,V.: Prosody Modeling for a Text-to-Speech
System by Artificial Neural Networks. Proc. of IASTED International
Conference "Signal and Image Processing 2000", November 19-23,2000,
Las Vegas, Nevada USA, pp.307-312, ISBN: 0-88986-308-3, ISNN:
1482-7921
[10] Sebesta,V., Tuckova,J.: Optimisation of Artificial Neural
Network Topology applied in the Prosody Control in Text-to-Speech
Synthesis. Theory and practice of informatics: Proc. of 27^th Annual
Conference on Current Trends in Theory and Practice of Informatics
"SOFSEM 2000", November 25-December 2, 2000, Milovy, Czech Republic,
pp.420-430, ISBN: 3-540-41348-0 Springer-Verlag Berlin Heidelberg New
York, ISSN: 0302-9743.
[11] Sebesta,V., Tuckova,J.: Application of Feature Extraction in
Text-to-Speech Processing.
Proc. of the International Conference on "Artificial Neural
Nets and Genetic Algorithms. Avril 2001, Prague, Czech Republic,
pp.145-148, ISBN: 3-211-83651-9 Springer-Verlag Wien-New York
[12] Tuckova,J., Sebesta, V.: Data Mining Approach for Prosody
Modelling by ANN in Text-to-Speech Synthesis. Proc. of the IASTED
Inernational Conference on "Artificial Intelligence and Applications -
AIA 2001", Ed. M.H.Hamza, ACTA Press Anaheim-Calgary-Zurich, ISBN:
0-88986-301-6, ISSN: 1482-7913, pp. 164-166, September 2001,
Marbella, Spain.
Chapter of the book
[13] Tuckova,J.,Bores,P.: The Neural Network Approach in Fundamental
Frequency Control. Speech Processing: Forum Phoneticum (Ed.:
Wodarz,H.-W.), Frankfurt am Main, Hector Verlag, 1997,pp. 143-154.
ISBN: 3-930220-10-5, ISSN: 0341-3144.
[14] Tuckova, J, Sebesta,V.: Optimalization of Input Data for
Fundamental Frequency Modelling by Neural Networks. Papers in
Phonetics and Speech Processing Forum Phoneticum 70 (2000),
pp.161-175. Ed. Hans-Walter Wodarz, ISBN: 3-930110-18-0 HECTOR -
Frankfurt am Main, ISSN: 0341-3144
International journals
[15] Sebesta,V., Tuckova, J.: Optimisation of Prosody Control in
Text-to-Speech Processing by Neural Networks. Int.Journal on Neural
and Mass-Parallel Computing and Information Systems "Neural Network
World", vol.9, no5, Prague 1999, pp.413-424
International workshops
[16 ] Tuckova,J., Horak,P.: Fundamental Frequency Control in Czech
Text-to-Speech Synthesis. Third Workshop on Electronic Control and
Measuring Systems, Université Paul Sabatier, Toulouse, France, June
1997, pp.80-84
[17] Tuckova J.: Investigation of Neural Net Pruning in F0 Modelling.
8-th Czech-German Workshop, Speech Processing, Faculty of Arts,
Charles University, Prague, September 1998. ISBN: 80-86269-00-0
[18] Tuckova, J., Strecha, G.: Automatic Labelling of Natural Speech
by Comparison with Synthetic Speech. 4^th International Workshop on
Electronics, Control, Measurement and Signals ECMS'99, Liberec, May
31-June 1, 1999, Czech Republic, pp.156-159.
[19] Tuckova, J.: Fundamental Frequency and Duration Modelling by
Neural Networks. 9^th Czech-German Workshop Speech Processing,
September 13-15, 1999, Prague, Czech Republic.
[20] Santarius,J., Tuckova,J.: Nnlab-Base of Platform for
Text-to-Speech Synthesis. 10^th Czech-German Workshop "Speech
Processing", September 18-20, 2000, Prague.
[29] Tuckova,J.: New Method of Input Data Creation for Prosody
Modelling. Proc. of 5-th Workshop on Electronics, Control, Modelling,
Measurement and Signals "ECM2S5 2001", p.279, Universite Paul
Sabatier, Toulouse, France, May 2001.
Proceedings of the national conferences, Lectures in Colloquium
[30] Tuckova J.: Influence of the Neural Net Input Parameters to F0
Training (In Czech). Koloquium No 595 - Circuit Theorie, Systems and
Signals. IREE, Academie of Sciences of Czech Republic, Prague, January
1998.
[31] Tuckova J.: Fundamental Frequency Control in TTS System.
Proc.of 8-th Int. Czech-Slovak Scientific Conference "Radioelektronika
'98, Brno, April 1998, Vol.1, pp.216-219
[32] Tuckova J.,Krkoska,M.,Musil,V.: A Neural Net Architecture in
Czech Text-to-Speech Synthesis. Proc.of 5-th Electronic Devices and
Systems Conference EDS'98, ISBN: 80-214-11-98-8, Brno, June 1998,
pp.359-361.
[33] Tuckova, J.: Comparison of the Czech and Swiss Mode of the F[0
]Modelling by Neural
Networks. (In Czech) Kolokvium No 606, IREE AS of Czech
Republic, Prague, 24 February 1999, Abstract on Internet pages of
IREE.
[34] Tuckova, J.: Influence of the ANN Canonical Structure on F[0
]Control in TTS Speech Synthesis. (In Czech). Seminar of the Circuit
Theory STO-7, Brno, September 16-17, 1999, Czech Republic, pp.188-191.
[35] Santarius,J., Tuckova,J.: Neural Net Laboratory. Proc. of the
Electronic Devices and
Systems Y2K. Ed. Vladislav Musil, VUT Brno, December 2000, ISBN
80-214-1780-3
[36] Santarius,J., Tuckova,J.: New Method of Text Pre-Processing for
Prosody Modelling. Proc.of 11-th Int. Czech-Slovak Scientific
Conference "Radioelektronika '2001, Brno, Czech Republic, May 2001
Vol.1, pp.154-157.
Invited lecture (In Czech)
[37] Sebesta,V., Tuckova, J.: Selection of Important Input Parameters
for a Text-to-Speech Synthesis by Neural Networks . "Hora
Informaticae" seminar, ICS AS CR, Prague, November 1999. Abstract on
Internet pages of ICS AS CR. Invited lecture.
[38] Tuckova, J.: Neural Networks and speech processing (In Czech).
Invited lecture for Universite of Zilina, Slovakia, Mars 23, 1999.
[39] Tuckova, J.: Neural Networks Applications .(In Czech). Invited
lecture, Department of telecommunication, Faculty of electrical
engineering, University of Zilina, Slovakia. 7.4.2000.
[40] Tuckova, J.: Task of the Neural Networks in the Speech Processing
(In Czech). Invited lecture for Universite of Zilina, Slovakia, Avril
6, 2001.
|
Update your profile