Research Associate Edvin Pakoci


Missing picture!

Edvin Pakoci

Research Associate


Telephone021/475-0204
E-mail
Academic titleResearch Associate
Employee's biography not found!
Research Associate
01.12.2013.
Historical data about the employee's membership not found!
TitleIn institution
Detekcija poplavljenog zemljišta na satelitskim slikama primenom Bajesovih klasifikatora

Bachelor's thesis

Electrical and Computer Engineering

Faculty of Technical Sciences

2011

Sinteza govora na bazi skrivenih Markovljevih modela za srpski jezik

Master's thesis

Electrical and Computer Engineering

Faculty of Technical Sciences

2012

Criterium of productionDescription
() Monographic study/chapter in M12 or research in the thematic collection of the international importanceSečujski M., Pekar D., Mak R., Pakoci E., Jakovljević N., Delić V., Popović B.: A comparison of decision tree based approaches to the modelling of prosodic features in Serbian, u: Verbal Communication Quality, Interdisciplinary Research II, 2013, Beograd: Centar za unapređenje životnih aktivnosti, pp. 75-90, ISBN 978-86-81879-46-7
() Research published in the international journalOstrogonac S., Pakoci E., Sečujski M., Mišković D.: Morphology-based vs Unsupervised Word Clustering for Training Language Models for Serbian, u: Acta Polytechnica Hungarica, Journal of Applied Sciences, 2018, Budimpešta: Univerzitet Obuda, ISSN 1785-8860, u štampi
() Report from the international meeting printed in its entiretyPopović B., Pakoci E., Pekar D.: A Comparison of Language Model Training Techniques in a Continuous Speech Recognition System for Serbian, Springer-Verlag, Lecture Notes in Artificial Intelligence, 2018, Vol. 11096, pp. 522-531, 20. konferencija Speech and Computer SPECOM, Lajpcig, 18-22 septembar, 2018, ISBN 978-3-319-99578-6
() Report from the international meeting printed in its entiretyPakoci E., Popović B., Pekar D.: Language Model Optimization for a Deep Neural Network Based Speech Recognition System for Serbian, Springer-Verlag, Lecture Notes in Artificial Intelligence, 2017, Vol. 10458, pp. 483-492, 19. konferencija Speech and Computer SPECOM, Hetfild, 12-16 septembar, 2017, ISBN 978-3-319-66428-6
() Report from the international meeting printed in its entiretyPakoci E., Popović B., Jakovljević N., Pekar D., Yassa F.: A Phonetic Segmentation Procedure Based on Hidden Markov Models, Springer-Verlag, Lecture Notes in Artificial Intelligence, 2016, Vol. 9811, pp. 67-74, 18. konferencija Speech and Computer SPECOM, Budimpešta, 23-27. avgust, 2016, ISBN 978-3-319-43958-7
() Report from the international meeting printed in its entiretyPopović B., Pakoci E., Pekar D.: Advanced Voice Activity Detection on Mobile Phones by Using Microphone Array and Phoneme-Specific Gaussian Mixture Models, 14. konferencija Symposium on Intelligent Systems and Informatics SISY, 2016, Subotica: IEEE, SMC Technical Committee on Computational Cybernetics, 29-31. avgust, 2016, pp. 45-48, ISBN: 978-1-5090-2866-5
() Report from the international meeting printed in its entiretyPopović B., Pakoci E., Jakovljević N., Kočiš G., Pekar D.: Voice Assistant Application for the Serbian Language, 23. Telekomunikacioni forum TELFOR, Beograd: IEEE, Telecommunications Society, 24-25. novembar, 2015, pp. 858-861, ISBN 978-1-5090-0054-8
() Report from the international meeting printed in its entiretyPakoci E., Jakovljević N., Popović B., Mišković D., Pekar D.: Speaker Detection Using Phoneme Specific Hidden Markov Models, Springer-Verlag, Lecture Notes in Computer Science, 2014, Vol. 8773, pp. 410-417, 16. konferencija Speech and Computer SPECOM, Novi Sad, 5-9. oktobar, 2014, ISBN 978-3-319-11580-1
() Report from the international meeting printed in its entiretyPopović B., Pakoci E., Pekar D.: End-to-End Large Vocabulary Speech Recognition for the Serbian Language, Springer-Verlag, Lecture Notes in Artificial Intelligence, 2017, Vol. 10458, pp. 343-352, 19. konferencija Speech and Computer SPECOM, Hetfild, 12-16 septembar, 2017, ISBN 978-3-319-66428-6
() Report from the international meeting printed in its entiretyPopović B., Ostrogonac S., Pakoci E., Jakovljević N., Delić V.: Deep Neural Network Based Continuous Speech Recognition for Serbian Using the Kaldi Toolkit, Springer-Verlag, Lecture Notes in Artificial Intelligence, 2015, Vol. 9319, pp. 186-192, 17. konferencija Speech and Computer SPECOM, Atina, 20-24. septembar, 2015, ISBN 978-3-319-23131-0
() Report from the international meeting printed in its entiretyPekar D., Mišković D., Suzić S., Pakoci E., Delić V.: Axon Voice Assistant - An Android-Based Smartphone Application in Serbian, 11. konferencija ETAI, Ohrid: Society for Electronics, Telecommunications, Automation and Informatics of the Republic of Macedonia, 26-28. septembar, 2013, pp. 1-6, ISBN 978-9989-630-68-2
() Report from the international meeting printed in its entiretyOstrogonac S., Suzić S., Bojanić M., Pakoci E.: Speech Resources for a Serbian LVCSR System, 21. Telekomunikacioni forum TELFOR, Beograd: IEEE, Telecommunications Society, 26-28. novembar, 2013, pp. 478-481, ISBN 978-1-4799-1419-7
() Report from the international meeting printed in its entiretyPakoci E., Mak R., Ostrogonac S.: Subjective Assessment of Text-to-Speech Synthesis Systems for the Serbian Language, 20. Telekomunikacioni forum TELFOR, Beograd: IEEE, Telecommunications Society, 20-22. novembar, 2012, pp. 732-735, ISBN 978-1-4673-2983-5
() Report from the international meeting printed in extractsPopović B., Pakoci E., Jakovljević N., Pekar D., Ostrogonac S.: A Robust Voice Activity Detection in Voice Assistant Application, 3. konferencija International Acoustics and Audio Engineering Conference TAKTONS, Novi Sad: RTV Vojvodine, AES, Dirigent Acoustics, Fakultet tehničkih nauka Novi Sad, 18-21. novembar, 2015, pp. 38-39, ISBN: 978-86-7892-758-4
() Research published in the journal of the national importancePakoci E., Popović B., Pekar D.: Improvements in Serbian Speech Recognition Using Sequence-Trained Deep Neural Networks, u: SPIIRAS Proceedings, 2018, Vol. 3, No. 58, pp. 53-76, ISSN 2078-9181
() Research published in the journal of the national importanceSuzić S., Ostrogonac S., Pakoci E., Bojanić M.: Building a Speech Repository for a Serbian LVCSR System, u: TELFOR Journal, 2014, Vol. 6, No. 2, pp. 109-114, ISSN 1821-3251
() Report from the meeting of the national importance printed in its entiretyPakoci E., Popović B., Pekar D.: Fast Sequence-Trained Deep Neural Network Models for Serbian Speech Recognition, 11. konferencija Digitalna obrada govora i slike DOGS, Novi Sad: Fakultet tehničkih nauka (Novi Sad), Elektrotehnički fakultet (Beograd), Elektronski fakultet (Niš), 22-25 novembar, 2017, pp. 25-28, ISBN 978-86-7892-993-9
() Report from the meeting of the national importance printed in its entiretyPopović B., Pakoci E., Ostrogonac S., Pekar D.: Large Vocabulary Continuous Speech Recognition for Serbian Using the Kaldi Toolkit, 10. konferencija Digitalna obrada govora i slike DOGS, Novi Sad: Fakultet tehničkih nauka (Novi Sad), Elektrotehnički fakultet (Beograd), Elektronski fakultet (Niš), 5-9. oktobar, 2014, pp. 31-34, ISBN 978-86-7892-633-4
() Report from the meeting of the national importance printed in its entiretyPopović B., Pakoci E., Jakovljević N., Pekar D.: Experiments With Regression Trees for Speaker Adaptation in Continuous Speech Recognition, 10. konferencija Digitalna obrada govora i slike DOGS, Novi Sad: Fakultet tehničkih nauka (Novi Sad), Elektrotehnički fakultet (Beograd), Elektronski fakultet (Niš), 5-9. oktobar, 2014, pp. 47-50, ISBN 978-86-7892-633-4
() Report from the meeting of the national importance printed in its entiretyPakoci E., Mak R.: HMM-Based Speech Synthesis for the Serbian Language, 56. Konferencija za elektroniku, telekomunikacije, računarstvo, automatiku i nuklearnu tehniku ETRAN, Zlatibor: Društvo za ETRAN, 11-14. jun, 2012, pp. TE4.7:1-4, ISBN 978-86-80509-67-9
() Report from the meeting of the national importance printed in its entiretyMak R., Pakoci E.: Use of Global Variance for the Enhancement of HMM-Based Speech Synthesis, 56. Konferencija za elektroniku, telekomunikacije, računarstvo, automatiku i nuklearnu tehniku ETRAN, Zlatibor: Društvo za ETRAN, 11-14. jun, 2012, pp. TE4.6:1-4, ISBN 978-86-80509-67-9
() New product or technology introduced in the production - international levelPakoci E., Popović B., Pekar D., Jakovljević N.: Skup modula i procedura za visokokvalitetnu anotaciju govornih baza podataka, http://www.ftn.uns.ac.rs/n1062251030/skup-modula-i-procedura-za-visokokvalitetnu-anotaciju-govornih-baza-podataka, korisnici: AlfaNum - govorne tehnologije d.o.o. (Novi Sad, Srbija), Speech Morphing System, Inc. (Campbell, CA, USA), 2015
() New product or technology introduced in the production - international levelMišković D., Pekar D., Sečujski M., Pakoci E.: SpeechLabel – alat za fonetsko i prozodijsko labeliranje govornih baza, http://www.ftn.uns.ac.rs/n1062251031/speechlabel---alat-za-fonetsko-i-prozodijsko-labeliranje-govornih-baza-podataka, korisnici: AlfaNum - govorne tehnologije d.o.o. (Novi Sad, Srbija), Speech Morphing System, Inc (Campbell, CA, USA), 2015
() New product line, new material, industrial prototypePopović B., Pakoci E., Pekar D., Delić V.: Napredna detekcija govornih segmenata na mobilnom telefonu primenom mikrofonskog niza i dekodera“, http://www.ftn.uns.ac.rs/n1871597686%/napredna-detekcija-govornih-segmenata-na-mobilnom-telefonu-primenom-mikrofonskog-niza-i-dekodera, korisnik: AlfaNum - govorne tehnologije d.o.o. (Novi Sad, Srbija), 2016
() New product line, new material, industrial prototypeMišković D., Vujnović Sedlar N., Ostrogonac S., Bojanić M., Mak R., Pakoci E.: Sistem za automatsko anketiranje (anDialer), http://www.ftn.uns.ac.rs/1740741199/sistem-za-automatsko-anketiranje--andialer-, korisnik: AlfaNum - govorne tehnologije d.o.o. (Novi Sad, Srbija), 2012
() Significantly improved existing product or technologyPopović B., Pakoci E., Pekar D., Delić V.: Poboljšanje performansi sistema za glasovno pozivanje (VoiceDial) primenom Chain dekodera, http://www.ftn.uns.ac.rs/995788631/poboljsanje-performansi-sistema-za-glasovno-pozivanje--voicedial--primenom-chain-dekodera-, korisnik: AlfaNum - govorne tehnologije d.o.o. (Novi Sad, Srbija), 2017
() Prototype, new methods, software, standardized and certified instrumentPopović B., Pakoci E., Jakovljević N., Mišković D., Suzić S., Delić V.: Axon govorni asistent - aplikacija za upravljanje mobilnim telefonom putem glasa na srpskom jeziku, http://www.ftn.uns.ac.rs/n4670117/axon-govorni-asistent---aplikacija-za-upravljanje-mobilnim-telefonom-putem-glasa-na-srpskom-jeziku, korisnik: AlfaNum - govorne tehnologije d.o.o. (Novi Sad, Srbija), 2015
() Prototype, new methods, software, standardized and certified instrumentPakoci E., Vujnović Sedlar, N., Ostrogonac S., Mak R.: Wavfilt – aplikacija za modifikovanje audio datoteka, http://www.ftn.uns.ac.rs/n4670115/wavfilt---aplikacija-za-modifikovanje-audio-datoteka, korisnik: AlfaNum - govorne tehnologije d.o.o. (Novi Sad, Srbija), 2015
() Prototype, new methods, software, standardized and certified instrumentSuzić S., Pekar D., Mak R., Pakoci E., Delić V.: Sistem za automatsko kreiranje transkribovane govorne baze (anSpeechCollector), http://www.ftn.uns.ac.rs/n4670235/sistem-za-automatsko-kreiranje-transkribovane-govorne-baze, korisnik: Fakultet tehničkih nauka i AlfaNum - govorne tehnologije d.o.o. (Novi Sad, Srbija), 2014
() Prototype, new methods, software, standardized and certified instrumentPakoci E., Mak R., Ostrogonac S., Jakovljević N., Sečujski M.: Sintetizator govora na bazi skrivenih Markovljevih modela za srpski jezik, http://www.ftn.uns.ac.rs/n1524171304/sintetizator-govora-na-bazi-skrivenih-markovljevih-modela-za-srpski-jezik, korisnik: AlfaNum - govorne tehnologije d.o.o. (Novi Sad, Srbija), 2012
() Prototype, new methods, software, standardized and certified instrumentJakovljević N., Mišković D., Pakoci E., Mak R., Popović B.: Dekoder za prepoznavanje kontinualnog govora na velikim rečnicima, http://www.ftn.uns.ac.rs/n1524171306/dekoder-za-prepoznavanje-kontinualnog-govora-na-velikim-recnicima, korisnik: AlfaNum - govorne tehnologije d.o.o. (Novi Sad, Srbija), 2012
() Realized patent, breed, sort or race, architectural, civil and urban original work Pekar D., Popović B., Pakoci E., Mišković D., Delić V., Sečujski M.: Duboke neuronske mreže na serveru u kombinaciji sa lokalnim dekoderom za prepoznavanje govora (Cloud-Based Deep Neural Networks Combined with Local Speech Recognition Decoder), AlfaNum - govorne tehnologije d.o.o. (Novi Sad, Srbija), 2018, broj prijave P-2016/0452, objavljen u Glasniku intelektualne svojine, broj 1/2018
TitleScientific areaHigher education institutionElection date
Research AssociateTelecommunications and Signal ProcessingFaculty of Technical Sciences27.11.2016.
Research AssociateTelecommunications and Signal ProcessingFaculty of Technical Sciences01.12.2013.