 |
Oytun Turk
OBJECTIVE:
To be employed as a post doctoral or senior research scientist in speech and
audio processing projects.
QUALIFICATIONS:
. Scientific research and technology development experience in speech and audio
processing with a focus on
voice conversion/modification, text-to-speech synthesis, acoustic
analysis/modeling, machine learning and
pattern recognition techniques for speech and audio processing.
. Experience in scientific research project management.
. Patent applications in voice conversion.
. Author of papers in speech/audio processing, speech therapy, and linguistics.
. PhD, MS, and BS degrees in Electrical and Electronics Engineering.
. Programming experience as a lead software engineer in complex programming
projects for text-to-speech synthesis,
voice conversion/modification, voice quality analysis, emotion research, and
computer-enabled speech therapy.
Working experience in C/C++, Java, Matlab, Visual C++ (MFC), Visual Basic;
experience in HTK, Prolog, PERL, OpenGL,
ATL/COM.
. Languages: Turkish (native), English (advanced), German (beginner).
EXPERIENCE:
. Research Scientist, DFKI, Berlin, Germany (May 2007 – Present)
- Researcher in multilingual, expressive, and adaptive text-to-speech
synthesis (Open Mary TTS: http://mary.dfki.de), emotional synthesis, voice
conversion/modification, and audio effects in IDEAS4GAMES and PAVOQUE
(http://mary.dfki.de/pavoque) projects.
- Proposal preparation for EU FP7 STREP “Contextually-Aware Information
Assistant – CAIA” that aims to combine research on context awareness
and on socially competent interaction, in order to endow machines with the
capability to initiate an interaction with a human user.
- Developing acoustic modeling and modification modules for emotional TTS
integrated with voice conversion.
- Performing research on robust prosody, emotion, and voice quality
modification and interpolation techniques for concatenative TTS.
- Performing research on modification of unit selection and HMM-based
synthesis output for modification and interpolation of expressive style,
speaker identity, and for post-processing with audio effects to create custom
voices.
. R&D Manager, Sestek Inc., Istanbul, Turkey (January 2003 – April 2007):
- Project manager and lead research scientist for “VOX - Voice Conversion
Software Development Project” with Voxonic Inc., NY, USA
(www.voxonic.com). Member of the team that developed “VOX”, the first
commercially available voice conversion system in the world (US patent pending,
No: 11/271,325).
- Technical lead for approximately 100 voice conversion projects in USA for
customers including various music and movie production companies. Projects
consisted of monolingual and cross-lingual voice transformations to various
languages (English, French, German, Spanish, Italian, Portuguese, Cuban,
Chinese, Japanese, Russian, Turkish, etc.) and to various celebrity voices.
- Proposal preparation for European Union projects “Computer-Aided Tools
for Speech Therapy – CATS”, and “Extending RTD Capacity for
Speech Processing Research in Turkey – ESPRES” with partners from
Greece, Italy, and Turkey.
- Proposal preparation and management of TUBITAK TIDEB Projects “Language
Independent Automatic Voice Conversion System” (Jan 2002-Dec 2003, budget
$ 360.000) and “Software Tools for Speech Therapy and Voice Quality
Monitoring” (Jan 2004-Dec 2006, budget $ 460.000).
- Proposal preparation for TUBITAK TEYDEB Project “New Generation Call
Center Integrated With Speech Processing Technology” (Jan 2007 – Dec
2009, budget $ 1.100.000).
. Senior Research Scientist, Sestek Inc. (January 2003 - April 2007):
- Managed a group of 3-5 junior researchers in voice conversion and speech
therapy projects.
- Developed algorithms for voice conversion: subband and context based voice
conversion, vocal tract transformation using ANNs, RBFNs, and SOMs, donor
selection for voice conversion, prosody modeling and transformation, alignment
using HMMs and the HTK.
- Developed algorithms for speech and voice quality modification: vocal tract
and prosody transplantation, voice quality interpolation for emotional TTS.
- Performed research on speech processing tools for speech therapy: vowel
recognition using GMMs, isolated word recognition using HMMs, voice quality
analysis, objective prosody match scoring.
- Performed collaborative research in TTS projects with FPMS (Belgium), DFKI
and T-Systems (Germany), ICP (France), and IVML (Greece).
- Database collection and subjective testing experience for voice conversion,
prosody modeling, and speech therapy.
- Author of journal and conference papers on voice conversion, modification,
speech processing for speech therapy, and linguistics. Attended major
conferences on signal processing. IEEE SIU 2004 Best Student Paper Award, 2nd
place.
. Senior Software Engineer, Sestek Inc. (January 2003 - April 2007):
- Lead developer of VoDi real-time voice modification software integrated with
SkypeTM/MSN MessengerTM:
http://www.sestek.com.tr/english/E_Commerce.asp?prg=vodi4
- Lead developer of Sestek Real-Time Song Recognition Module, Sestek Real-Time
Voice Quality Analysis Module (as featured in MusicTutor 1.0), and an automatic
segmentation module (as featured in EasyTranscriber 1.0)
- Developed speech processing algorithms and software tools for speech therapy.
. Research Scientist, Sestek Inc. (September 2000 - December 2002):
- Performed scientific research on voice conversion and voice modification.
- Developed a basic GMM based speaker identification system in C/C++.
- Performed research on subband decomposition and wavelets for voice
conversion applications.
. Software Engineer, Sestek Inc. (September 2000 - December 2002):
- Developed a waveform analysis and editing interface for the voice conversion
software VOX in Visual C++ with the software development team.
- Developed objective and subjective testing interfaces for voice conversion.
- Lead developer of VowelRecognizer, a real-time Turkish vowel recognition
software for children.
- Implemented various speech/audio processing, machine learning and pattern
recognition algorithms in C/C++ and Matlab: HMMs/GMMs/Neural Networks,
clustering, PSOLA, vocal tract scaling, pitch tracking, end-point detection,
audio effects.
- Developed software tools for audio and electro-glottograph recording, audio
format conversion, filter design and filtering, noise reduction, batch
processing.
. Research Scientist, Bogazici University Multimedia Laboratory - BUMM (January
2003 – April 2007):
- Performed research on voice quality analysis and modification, and on speech
processing algorithms for speech therapy in the European Union funded Network
of Excellence SIMILAR (Ref. 507609, www.similar.cc).
- Performed research on the development of voice conversion techniques for the
assessment of speaker identification systems in the European Union funded
Network of Excellence project BIOSECURE (Ref. 507634).
. Visiting Researcher, Faculté Polytechnique de Mons, TCTS Laboratory,
Belgium (August 2004):
- Developed a new technique for voice quality interpolation in emotional TTS.
. Board Member, Sestek Inc., Istanbul, Turkey (September 2000 – April 2007)
- Start-up company in speech, audio and IVR technologies.
EDUCATION
. 2003-2007, PhD in Electrical and Electronics Engineering, Bogazici
University, Istanbul, Turkey
PhD Thesis: Cross-Lingual Voice Conversion (Developed new algorithms to
improve voice conversion performance and robustness in cross-lingual voice
conversion applications)
Advisor: Prof. Dr. Levent Arslan
Major Courses: Pattern Recognition, Artificial Neural Networks, Information
Theory, Detection/Estimation Theory, Speech Recognition, Digital Control.
GPA: 3.82/4.00
. 2000-2003, MS in Electrical and Electronics Engineering, Bogazici University,
Istanbul, Turkey
MS Thesis: New Methods for Voice Conversion (Developed algorithms, objective
and subjective evaluation methods for voice conversion)
Advisor: Assoc. Prof. Dr. Levent Arslan
Major Courses: Psycho-acoustics, Adaptive Filter Theory, Independent Component
Analysis, Machine Learning, Natural Language Processing, Spectral Analysis,
Statistical Signal Processing, Multiresolution Signal Processing.
GPA: 3.56/4.00
. 1995-2000, BS in Electrical and Electronics Engineering, Bogazici University,
Istanbul, Turkey
Final project: Automatic Polyphonic Music Transcription
Advisor: Assist. Prof. Dr. Levent Arslan
Major Courses: Signals and Systems, Digital Signal Processing, Speech
Processing, Image Processing, C/C++ and Java Programming, undergraduate courses
in systems and control.
. 1991-1994, Ankara Fen Lisesi, Ankara, Turkey.
LIST OF PATENTS & PUBLICATIONS:
. Publications available online:
http://www.sestek.com.tr/english/Products.asp?proId=21
http://www.dfki.de/~otuerk/
. Patents:
1.Türk, O., L. M. Arslan, and F. Deutsch. “Codebook-less Speech
Conversion Method and System”, Voxonic Inc., United States Patent
20070213987, Publication Date: 13/09/2007.
2.Türk, O., L. M. Arslan, and F. Deutsch. “Donor Selection for Voice
Conversion”, Voxonic Inc., Provisional U.S. Patent Application, 14 Mar
2005.
. Journals:
1.Türk, O., and Arslan, L., M., 2007, “Donor selection for Voice
Conversion”, Journal of the Acoust. Soc. Of America, submitted for review.
2.Türk, O., and Arslan, L. M., 2006, "Robust Processing Techniques for
Voice Conversion", Computer Speech and Language 20 (2006), pp. 441-467.
3.Sari, E., Ucar, C., Türk, O., Kurtulmus, H., Altug, H. A., Pocan, S.,
2008, “Treatment of a Patient with Cleft Lip and Palate Using an Internal
Distraction Device”, The Cleft Palate-Craniofacial Journal, accepted 14
January 2008.
. Conference Proceedings:
1.Türk, O., and Schröder, M., “A Comparison of Voice Conversion
Methods for Transforming Voice Quality in Emotional Speech Synthesis”,
submitted to Interspeech 2008, Brisbane, Australia.
2.Pammi, S. C., Charfuelan, M., Schröder, M., and Türk, O.,
“Voice Building Tool for the MARY TTS Platform”, submitted to
Interspeech 2008, Brisbane, Australia.
3.Schröder M., Gebhard, P., Charfuelan, M., Endres, C., Kipp, M., Pammi,
S., Rumpler, M., and Türk, O., “Enhancing Animated Agents in an
Instrumented Poker Game”, accepted for 31st German Conference on
Artificial Intelligence (KI 2008).
4.Gebhard, P., Schröder, M., Charfuelan, M., Endres, C., Kipp, M., Pammi,
S., Rumpler, M., and Türk, O., “IDEAS4Games: Building Expressive ECAs
for Computer Games”, submitted to the 8th International Conference on
Intelligent Virtual Agents (IVA 2008), Tokyo, Japan.
5.Burkhardt, F., Audibert, N., Malatesta, L., Türk, O., Arslan, L., and
Auberge, V. “Emotional Prosody - Does Culture Make A Difference?,”
Speech Prosody 2006, Dresden, Germany.
6.Türk, O., and Arslan, L. M., 2005, "Software Tools for Speech Therapy
and Voice Quality Monitoring", 13th European Signal Processing Conference -
EUSIPCO 2005, Antalya, Turkey.
7.Türk, O., and Arslan, L. M., 2005, "Donor Selection for Voice
Conversion", 13th European Signal Processing Conference - EUSIPCO 2005,
Antalya, Turkey.
8.Türk, O., Schröder, M., Bozkurt, B., and Arslan L. M., 2005, "Voice
Quality Interpolation for Emotional Text-To-Speech Synthesis", INTERSPEECH
2005, Lisbon, Portugal.,
9.Türk, O., and Arslan, L. M., 2004, "Pronunciation Scoring for the
Hearing-Impaired", in Proc. of the 9th International Conference Speech and
Computer - SPECOM 2004, St. Petersburg, Russia.
10.Türk, O., and Arslan, L. M., 2004, “Konusma Terapisine Yonelik
Otomatik Konusma Tanima Yontemleri”, II. Ulusal Dil ve Konusma
Bozukluklari Kongresi 2004, Eskisehir, Turkey.
11.Türk, O., Sayli, O., Ozsoy, A., S., and Arslan, L., M.,
“Türkce'de Ünlülerin Formant Frekans Incelemesi”,
18. Ulusal Dilbilim Kurultayi 2004, Ankara, Turkey.
12.Türk, O., and Arslan, L. M., 2004, “Dayanikli Konusmaci Donusturme
Yontemleri”, SIU 2004, Kusadasi, Turkey.
13.Türk, O., and Arslan, L. M., 2004, “Konusma Terapisine Yonelik
Konusma Tanima Yontemleri”, SIU 2004, Kusadasi, Turkey.
14.Türk, O., and Arslan, L. M., 2003, “New Methods for Vocal Tract
and Pitch Contour Transformation”, EUROSPEECH 2003, Geneva, Switzerland.
15.Türk, O., and Arslan, L. M., 2003, “Subjective Evaluations for
Perception of Speaker Identity Through Acoustic Feature Transplantations”,
EUROSPEECH 2003, Geneva, Switzerland.
16.Türk, O., and Arslan, L. M., 2003, “Konusmaci Donusturme Icin Uc
Yeni Yontem”, Proceedings of SIU 2003, pp. 398-401, Istanbul, Turkey.
17.Türk, O., and Arslan, L. M., 2003, “Konusmaci Kimligi
Algilanmasinda Akustik Ozniteliklerin Karsilastirmali Analizi”,
Proceedings of SIU 2003, pp. 394-397, Istanbul, Turkey.
18.Türk, O., and Arslan, L. M., 2002, “Subband Based Voice
Conversion”, Proceedings of the ICSLP 2002, vol. 1, pp. 289-292, September
2002, Denver, Colorado, USA.
19.Ormanci, E., Nikbay, U. H., Türk, O., and Arslan, L. M., 2002,
“Subjective Assessment of Frequency Bands for Perception of Speaker
Identity”, Proceedings of the ICSLP 2002, vol 4, pp. 2581-2584, September
2002, Denver, Colorado, USA.
20.Türk, O., Sayli, O., Dutagaci, H., and Arslan, L. M., 2002, “A
Sound Source Classification System Based On Subband Processing”, Proc. of
the 3rd IEEE Benelux Signal Processing Symposium (SPS-2002), Leuven, Belgium.
21.Türk, O., Sayli, O., Dutagaci, H., and Arslan, L. M., 2002,
“Alt-bant Islemeye Dayali Bir Ses Siniflandirma Sistemi”, SIU 2002,
Denizli, Turkey.
. Theses:
1.Türk, O., 2007 “Cross-Lingual Voice Conversion”, PhD Thesis,
Bogazici University.
2.Türk, O., 2003, “New Methods For Voice Conversion”, MS Thesis,
Bogazici University.
3.Türk, O., 2000, “Automated transcription methods for polyphonic
music signals,” BS project, Bogazici University.
REFERENCES
. Dr. Marc Schröder, Senior Researcher, DFKI GmbH Language Technology
Laboratory.
. Prof. Dr. Levent Arslan, Prof. of Electrical and Electronics Eng., Bogazici
University, Istanbul, Turkey; Board Member and CEO, Sestek Inc., Istanbul,
Turkey; former member of Entropic Research, Washington DC, US.
CERTIFICATES
. C and System Programmers Association, Turkey, C++ Certificate.
MEMBERSHIPS & HONORS
. IEEE Member since 2001.
. Reviewer for IEEE Transactions on Audio, Speech and Language Processing, IEEE
Signal Processing Magazine, IEEE Transactions on Computers.
|
Update your profile