Advanced search options

Advanced Search Options 🞨

Browse by author name (“Author name starts with…”).

Find ETDs with:

in
/  
in
/  
in
/  
in

Written in Published in Earliest date Latest date

Sorted by

Results per page:

Sorted by: relevance · author · university · dateNew search

You searched for subject:(speaker verification). Showing records 1 – 30 of 54 total matches.

[1] [2]

Search Limiters

Last 2 Years | English Only

Country

▼ Search Limiters


Universidade do Porto

1. Moura, Paulo André Alves. Controlo remoto de presenças recorrendo à tecnologia de speaker verification.

Degree: 2010, Universidade do Porto

Estágio realizado na PT Inovação e orientado pelo Eng.º Sérgio Ramalho

Tese de mestrado integrado. Engenharia Electrotécnica e de Computadores (Major Telecomunicações). Faculdade de Engenharia. Universidade do Porto. 2010

Advisors/Committee Members: Ferreira, Aníbal João de Sousa, Universidade do Porto. Faculdade de Engenharia.

Subjects/Keywords: Speaker verification

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Moura, P. A. A. (2010). Controlo remoto de presenças recorrendo à tecnologia de speaker verification. (Thesis). Universidade do Porto. Retrieved from http://www.rcaap.pt/detail.jsp?id=oai:repositorio-aberto.up.pt:10216/57628

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Moura, Paulo André Alves. “Controlo remoto de presenças recorrendo à tecnologia de speaker verification.” 2010. Thesis, Universidade do Porto. Accessed January 21, 2020. http://www.rcaap.pt/detail.jsp?id=oai:repositorio-aberto.up.pt:10216/57628.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Moura, Paulo André Alves. “Controlo remoto de presenças recorrendo à tecnologia de speaker verification.” 2010. Web. 21 Jan 2020.

Vancouver:

Moura PAA. Controlo remoto de presenças recorrendo à tecnologia de speaker verification. [Internet] [Thesis]. Universidade do Porto; 2010. [cited 2020 Jan 21]. Available from: http://www.rcaap.pt/detail.jsp?id=oai:repositorio-aberto.up.pt:10216/57628.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Moura PAA. Controlo remoto de presenças recorrendo à tecnologia de speaker verification. [Thesis]. Universidade do Porto; 2010. Available from: http://www.rcaap.pt/detail.jsp?id=oai:repositorio-aberto.up.pt:10216/57628

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


Clemson University

2. Afnan, Shamama. Comparison GMM and SVM Classifier for Automatic Speaker Verification.

Degree: MS, Electrical Engineering, 2015, Clemson University

  The objective of this thesis is to develop automatic text-independent speaker verification systems using unconstrained telephone conversational speech. We began by performing a Gaussian… (more)

Subjects/Keywords: Digital Signal Processing; GMM in Speaker Verification; Speaker Verification; SVM in Speaker Verification; Engineering

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Afnan, S. (2015). Comparison GMM and SVM Classifier for Automatic Speaker Verification. (Masters Thesis). Clemson University. Retrieved from https://tigerprints.clemson.edu/all_theses/2228

Chicago Manual of Style (16th Edition):

Afnan, Shamama. “Comparison GMM and SVM Classifier for Automatic Speaker Verification.” 2015. Masters Thesis, Clemson University. Accessed January 21, 2020. https://tigerprints.clemson.edu/all_theses/2228.

MLA Handbook (7th Edition):

Afnan, Shamama. “Comparison GMM and SVM Classifier for Automatic Speaker Verification.” 2015. Web. 21 Jan 2020.

Vancouver:

Afnan S. Comparison GMM and SVM Classifier for Automatic Speaker Verification. [Internet] [Masters thesis]. Clemson University; 2015. [cited 2020 Jan 21]. Available from: https://tigerprints.clemson.edu/all_theses/2228.

Council of Science Editors:

Afnan S. Comparison GMM and SVM Classifier for Automatic Speaker Verification. [Masters Thesis]. Clemson University; 2015. Available from: https://tigerprints.clemson.edu/all_theses/2228

3. Ambika D. Enhanced Approaches to Improve the Performance of Secure Speech Communication for Wireless Network;.

Degree: Computer Science, 2015, Avinashilingam Deemed University For Women

The tremendous growth envisaged in computer hardware software and network technologies have made communication in an easy and cost effective manner Recent newlinedevelopments in the… (more)

Subjects/Keywords: Speaker verification; speech water marking; scrambling

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

D, A. (2015). Enhanced Approaches to Improve the Performance of Secure Speech Communication for Wireless Network;. (Thesis). Avinashilingam Deemed University For Women. Retrieved from http://shodhganga.inflibnet.ac.in/handle/10603/54380

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

D, Ambika. “Enhanced Approaches to Improve the Performance of Secure Speech Communication for Wireless Network;.” 2015. Thesis, Avinashilingam Deemed University For Women. Accessed January 21, 2020. http://shodhganga.inflibnet.ac.in/handle/10603/54380.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

D, Ambika. “Enhanced Approaches to Improve the Performance of Secure Speech Communication for Wireless Network;.” 2015. Web. 21 Jan 2020.

Vancouver:

D A. Enhanced Approaches to Improve the Performance of Secure Speech Communication for Wireless Network;. [Internet] [Thesis]. Avinashilingam Deemed University For Women; 2015. [cited 2020 Jan 21]. Available from: http://shodhganga.inflibnet.ac.in/handle/10603/54380.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

D A. Enhanced Approaches to Improve the Performance of Secure Speech Communication for Wireless Network;. [Thesis]. Avinashilingam Deemed University For Women; 2015. Available from: http://shodhganga.inflibnet.ac.in/handle/10603/54380

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


Clemson University

4. Fang, Eric. Phoneme Weighting and Energy-Based Weighting for Speaker Recognition.

Degree: PhD, Computer Engineering, 2012, Clemson University

 This dissertation focuses on determining specific vowel phonemes which work best for speaker identification and speaker verification, and also developing new algorithms to improve speaker(more)

Subjects/Keywords: Gaussian mixture model; speaker identification; speaker recognition; speaker verification; vector quantization; Computer Engineering

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Fang, E. (2012). Phoneme Weighting and Energy-Based Weighting for Speaker Recognition. (Doctoral Dissertation). Clemson University. Retrieved from https://tigerprints.clemson.edu/all_dissertations/1063

Chicago Manual of Style (16th Edition):

Fang, Eric. “Phoneme Weighting and Energy-Based Weighting for Speaker Recognition.” 2012. Doctoral Dissertation, Clemson University. Accessed January 21, 2020. https://tigerprints.clemson.edu/all_dissertations/1063.

MLA Handbook (7th Edition):

Fang, Eric. “Phoneme Weighting and Energy-Based Weighting for Speaker Recognition.” 2012. Web. 21 Jan 2020.

Vancouver:

Fang E. Phoneme Weighting and Energy-Based Weighting for Speaker Recognition. [Internet] [Doctoral dissertation]. Clemson University; 2012. [cited 2020 Jan 21]. Available from: https://tigerprints.clemson.edu/all_dissertations/1063.

Council of Science Editors:

Fang E. Phoneme Weighting and Energy-Based Weighting for Speaker Recognition. [Doctoral Dissertation]. Clemson University; 2012. Available from: https://tigerprints.clemson.edu/all_dissertations/1063


Queensland University of Technology

5. Baker, Brendan J. Speaker verification incorporating high-level linguistic features.

Degree: 2008, Queensland University of Technology

Speaker verification is the process of verifying or disputing the claimed identity of a speaker based on a recorded sample of their speech. Automatic speaker(more)

Subjects/Keywords: speaker recognition; speaker verification; high-level features; idiolect; phonetic speaker recognition; session variability; text-constrained speaker recognition

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Baker, B. J. (2008). Speaker verification incorporating high-level linguistic features. (Thesis). Queensland University of Technology. Retrieved from http://eprints.qut.edu.au/17665/

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Baker, Brendan J. “Speaker verification incorporating high-level linguistic features.” 2008. Thesis, Queensland University of Technology. Accessed January 21, 2020. http://eprints.qut.edu.au/17665/.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Baker, Brendan J. “Speaker verification incorporating high-level linguistic features.” 2008. Web. 21 Jan 2020.

Vancouver:

Baker BJ. Speaker verification incorporating high-level linguistic features. [Internet] [Thesis]. Queensland University of Technology; 2008. [cited 2020 Jan 21]. Available from: http://eprints.qut.edu.au/17665/.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Baker BJ. Speaker verification incorporating high-level linguistic features. [Thesis]. Queensland University of Technology; 2008. Available from: http://eprints.qut.edu.au/17665/

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


University of Illinois – Urbana-Champaign

6. Pekcan, Onur. Development of machine learning based speaker recognition system.

Degree: MS, 0112, 2010, University of Illinois – Urbana-Champaign

 In this thesis, we describe a biometric authentication system that is capable of recognizing its users??? voice using advanced machine learning and digital signal processing… (more)

Subjects/Keywords: Speaker Recognition; Speaker Verification; Speaker Identification; Biometrics; Voice Activity Detection; Machine Learning; Support Vector Machine; Artificial Neural Network; Hardware; Software

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Pekcan, O. (2010). Development of machine learning based speaker recognition system. (Thesis). University of Illinois – Urbana-Champaign. Retrieved from http://hdl.handle.net/2142/16226

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Pekcan, Onur. “Development of machine learning based speaker recognition system.” 2010. Thesis, University of Illinois – Urbana-Champaign. Accessed January 21, 2020. http://hdl.handle.net/2142/16226.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Pekcan, Onur. “Development of machine learning based speaker recognition system.” 2010. Web. 21 Jan 2020.

Vancouver:

Pekcan O. Development of machine learning based speaker recognition system. [Internet] [Thesis]. University of Illinois – Urbana-Champaign; 2010. [cited 2020 Jan 21]. Available from: http://hdl.handle.net/2142/16226.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Pekcan O. Development of machine learning based speaker recognition system. [Thesis]. University of Illinois – Urbana-Champaign; 2010. Available from: http://hdl.handle.net/2142/16226

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


University of New South Wales

7. Sriskandaraja, Kaavya. Spoofing countermeasures for secure and robust voice authentication system: Feature extraction and modelling.

Degree: Electrical Engineering & Telecommunications, 2018, University of New South Wales

 The ability to employ automatic speaker verification systems without face-to-face contact makes them more prone to spoofing attacks compared to other biometric systems. The study… (more)

Subjects/Keywords: Spoofing countermeasures; Speech Processing; Speaker verification; Voice authentication

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Sriskandaraja, K. (2018). Spoofing countermeasures for secure and robust voice authentication system: Feature extraction and modelling. (Doctoral Dissertation). University of New South Wales. Retrieved from http://handle.unsw.edu.au/1959.4/60356 ; https://unsworks.unsw.edu.au/fapi/datastream/unsworks:51915/SOURCE02?view=true

Chicago Manual of Style (16th Edition):

Sriskandaraja, Kaavya. “Spoofing countermeasures for secure and robust voice authentication system: Feature extraction and modelling.” 2018. Doctoral Dissertation, University of New South Wales. Accessed January 21, 2020. http://handle.unsw.edu.au/1959.4/60356 ; https://unsworks.unsw.edu.au/fapi/datastream/unsworks:51915/SOURCE02?view=true.

MLA Handbook (7th Edition):

Sriskandaraja, Kaavya. “Spoofing countermeasures for secure and robust voice authentication system: Feature extraction and modelling.” 2018. Web. 21 Jan 2020.

Vancouver:

Sriskandaraja K. Spoofing countermeasures for secure and robust voice authentication system: Feature extraction and modelling. [Internet] [Doctoral dissertation]. University of New South Wales; 2018. [cited 2020 Jan 21]. Available from: http://handle.unsw.edu.au/1959.4/60356 ; https://unsworks.unsw.edu.au/fapi/datastream/unsworks:51915/SOURCE02?view=true.

Council of Science Editors:

Sriskandaraja K. Spoofing countermeasures for secure and robust voice authentication system: Feature extraction and modelling. [Doctoral Dissertation]. University of New South Wales; 2018. Available from: http://handle.unsw.edu.au/1959.4/60356 ; https://unsworks.unsw.edu.au/fapi/datastream/unsworks:51915/SOURCE02?view=true

8. Finan, Robert Andrew. Towards the use of sub-band processing in automatic speaker recognition.

Degree: PhD, 1998, Abertay University

 Automatic speaker recognition uses a person's voice as a means of identifying them. It has many practical applications, particularly in the area of security systems.… (more)

Subjects/Keywords: 621.3994; Speaker verification; Score normalisation

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Finan, R. A. (1998). Towards the use of sub-band processing in automatic speaker recognition. (Doctoral Dissertation). Abertay University. Retrieved from https://rke.abertay.ac.uk/en/studentTheses/93e64b07-705b-4ffe-a42b-1d31e69c59b8 ; https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.263686

Chicago Manual of Style (16th Edition):

Finan, Robert Andrew. “Towards the use of sub-band processing in automatic speaker recognition.” 1998. Doctoral Dissertation, Abertay University. Accessed January 21, 2020. https://rke.abertay.ac.uk/en/studentTheses/93e64b07-705b-4ffe-a42b-1d31e69c59b8 ; https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.263686.

MLA Handbook (7th Edition):

Finan, Robert Andrew. “Towards the use of sub-band processing in automatic speaker recognition.” 1998. Web. 21 Jan 2020.

Vancouver:

Finan RA. Towards the use of sub-band processing in automatic speaker recognition. [Internet] [Doctoral dissertation]. Abertay University; 1998. [cited 2020 Jan 21]. Available from: https://rke.abertay.ac.uk/en/studentTheses/93e64b07-705b-4ffe-a42b-1d31e69c59b8 ; https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.263686.

Council of Science Editors:

Finan RA. Towards the use of sub-band processing in automatic speaker recognition. [Doctoral Dissertation]. Abertay University; 1998. Available from: https://rke.abertay.ac.uk/en/studentTheses/93e64b07-705b-4ffe-a42b-1d31e69c59b8 ; https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.263686


Queensland University of Technology

9. McLaren, Mitchell Leigh. Improving automatic speaker verification using SVM techniques.

Degree: 2009, Queensland University of Technology

 Automatic recognition of people is an active field of research with important forensic and security applications. In these applications, it is not always possible for… (more)

Subjects/Keywords: speaker recognition; speaker verification; support vector machines; Gaussian mixture models; session variation; unsupervised adaptation; data selection

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

McLaren, M. L. (2009). Improving automatic speaker verification using SVM techniques. (Thesis). Queensland University of Technology. Retrieved from https://eprints.qut.edu.au/32063/

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

McLaren, Mitchell Leigh. “Improving automatic speaker verification using SVM techniques.” 2009. Thesis, Queensland University of Technology. Accessed January 21, 2020. https://eprints.qut.edu.au/32063/.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

McLaren, Mitchell Leigh. “Improving automatic speaker verification using SVM techniques.” 2009. Web. 21 Jan 2020.

Vancouver:

McLaren ML. Improving automatic speaker verification using SVM techniques. [Internet] [Thesis]. Queensland University of Technology; 2009. [cited 2020 Jan 21]. Available from: https://eprints.qut.edu.au/32063/.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

McLaren ML. Improving automatic speaker verification using SVM techniques. [Thesis]. Queensland University of Technology; 2009. Available from: https://eprints.qut.edu.au/32063/

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


Brno University of Technology

10. Kaňok, Tomáš. Rozpoznávání mluvčího ve Skype hovorech .

Degree: 2011, Brno University of Technology

 Tato diplomová práce se zabývá problematikou strojové identifikace a verifikace řečníka, její teorií a aplikací. Vyhodnocuje existující implementaci dané problematiky skupinou [email protected] Dále se zabývá… (more)

Subjects/Keywords: verifikace řečníka; identifikace rečníka; SkypeAPI; zásuvný modul; BSAPI; [email protected]; Qt; speaker verification; speaker identification; SkypeAPI; plugin; BSAPI; [email protected]; Qt

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Kaňok, T. (2011). Rozpoznávání mluvčího ve Skype hovorech . (Thesis). Brno University of Technology. Retrieved from http://hdl.handle.net/11012/52761

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Kaňok, Tomáš. “Rozpoznávání mluvčího ve Skype hovorech .” 2011. Thesis, Brno University of Technology. Accessed January 21, 2020. http://hdl.handle.net/11012/52761.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Kaňok, Tomáš. “Rozpoznávání mluvčího ve Skype hovorech .” 2011. Web. 21 Jan 2020.

Vancouver:

Kaňok T. Rozpoznávání mluvčího ve Skype hovorech . [Internet] [Thesis]. Brno University of Technology; 2011. [cited 2020 Jan 21]. Available from: http://hdl.handle.net/11012/52761.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Kaňok T. Rozpoznávání mluvčího ve Skype hovorech . [Thesis]. Brno University of Technology; 2011. Available from: http://hdl.handle.net/11012/52761

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


Brno University of Technology

11. Profant, Ján. Robustní rozpoznávání mluvčího pomocí neuronových sítí .

Degree: 2019, Brno University of Technology

 Tématem této práce je analýza nejmodernějších systémů pro rozpoznávání řečníka za použití neurónových sítí (nazývaných x-vektory) v rozličných podmínkách, jako jsou širokopásmové a úzkopásmové data,… (more)

Subjects/Keywords: verifikace mluvčího; rozpoznávání mluvčího; neurónové sítě; x-vector; i-vector; speaker verification; speaker recognition; neural networks; x-vector; i-vector

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Profant, J. (2019). Robustní rozpoznávání mluvčího pomocí neuronových sítí . (Thesis). Brno University of Technology. Retrieved from http://hdl.handle.net/11012/180393

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Profant, Ján. “Robustní rozpoznávání mluvčího pomocí neuronových sítí .” 2019. Thesis, Brno University of Technology. Accessed January 21, 2020. http://hdl.handle.net/11012/180393.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Profant, Ján. “Robustní rozpoznávání mluvčího pomocí neuronových sítí .” 2019. Web. 21 Jan 2020.

Vancouver:

Profant J. Robustní rozpoznávání mluvčího pomocí neuronových sítí . [Internet] [Thesis]. Brno University of Technology; 2019. [cited 2020 Jan 21]. Available from: http://hdl.handle.net/11012/180393.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Profant J. Robustní rozpoznávání mluvčího pomocí neuronových sítí . [Thesis]. Brno University of Technology; 2019. Available from: http://hdl.handle.net/11012/180393

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


Brno University of Technology

12. Kašpar, Ladislav. Rozpoznávání mluvčího .

Degree: 2013, Brno University of Technology

 Ve své bakalářské práci se věnuji problému rozpoznávání mluvčího. Tato práce obsahuje základní teorie k tomuto tématu. Teorie je zaměřena na výpočet parametrů pro rozpoznávání… (more)

Subjects/Keywords: Rozpoznávání mluvčího; verifikace mluvčího; identifikace mluvčího; vytváření řeči; segmentace; analýza řečového signálu; formanty.; Speaker recognition; speaker verification; speaker identification; speech production; segmentation; analysis of the speech signal; formants.

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Kašpar, L. (2013). Rozpoznávání mluvčího . (Thesis). Brno University of Technology. Retrieved from http://hdl.handle.net/11012/27331

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Kašpar, Ladislav. “Rozpoznávání mluvčího .” 2013. Thesis, Brno University of Technology. Accessed January 21, 2020. http://hdl.handle.net/11012/27331.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Kašpar, Ladislav. “Rozpoznávání mluvčího .” 2013. Web. 21 Jan 2020.

Vancouver:

Kašpar L. Rozpoznávání mluvčího . [Internet] [Thesis]. Brno University of Technology; 2013. [cited 2020 Jan 21]. Available from: http://hdl.handle.net/11012/27331.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Kašpar L. Rozpoznávání mluvčího . [Thesis]. Brno University of Technology; 2013. Available from: http://hdl.handle.net/11012/27331

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


Brno University of Technology

13. Pešán, Jan. Rozpoznávání mluvčího na mobilním telefonu .

Degree: 2011, Brno University of Technology

 Tato práce se zaměřuje na implementaci počítačového systému rozpoznávání řečníka do prostředí mobilního telefonu. Je zde popsán princip, funkce, a implementace rozpoznávače na mobilním telefonu… (more)

Subjects/Keywords: Rozpoznání řečníka; Identifikace řečníka; Ověření řečníka; Nokia N900; ARM; iVector; PLDA; Biometrie; Speaker recognition; Speaker identification; Speaker verification; Nokia N900; ARM; iVector; PLDA; Biometry

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Pešán, J. (2011). Rozpoznávání mluvčího na mobilním telefonu . (Thesis). Brno University of Technology. Retrieved from http://hdl.handle.net/11012/54199

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Pešán, Jan. “Rozpoznávání mluvčího na mobilním telefonu .” 2011. Thesis, Brno University of Technology. Accessed January 21, 2020. http://hdl.handle.net/11012/54199.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Pešán, Jan. “Rozpoznávání mluvčího na mobilním telefonu .” 2011. Web. 21 Jan 2020.

Vancouver:

Pešán J. Rozpoznávání mluvčího na mobilním telefonu . [Internet] [Thesis]. Brno University of Technology; 2011. [cited 2020 Jan 21]. Available from: http://hdl.handle.net/11012/54199.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Pešán J. Rozpoznávání mluvčího na mobilním telefonu . [Thesis]. Brno University of Technology; 2011. Available from: http://hdl.handle.net/11012/54199

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

14. Remya, A R. Efficient Watermarking Schemes for Speaker Verification Guaranteeing Non-repudiation.

Degree: 2014, Cochin University of Science and Technology

Presently different audio watermarking methods are available; most of them inclined towards copyright protection and copy protection. This is the key motive for the notion… (more)

Subjects/Keywords: Signal Processing; Watermarking; Speaker verification; Barcode Based FeatureMarking Scheme; Human Auditory System

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Remya, A. R. (2014). Efficient Watermarking Schemes for Speaker Verification Guaranteeing Non-repudiation. (Thesis). Cochin University of Science and Technology. Retrieved from http://dyuthi.cusat.ac.in/purl/4959

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Remya, A R. “Efficient Watermarking Schemes for Speaker Verification Guaranteeing Non-repudiation.” 2014. Thesis, Cochin University of Science and Technology. Accessed January 21, 2020. http://dyuthi.cusat.ac.in/purl/4959.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Remya, A R. “Efficient Watermarking Schemes for Speaker Verification Guaranteeing Non-repudiation.” 2014. Web. 21 Jan 2020.

Vancouver:

Remya AR. Efficient Watermarking Schemes for Speaker Verification Guaranteeing Non-repudiation. [Internet] [Thesis]. Cochin University of Science and Technology; 2014. [cited 2020 Jan 21]. Available from: http://dyuthi.cusat.ac.in/purl/4959.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Remya AR. Efficient Watermarking Schemes for Speaker Verification Guaranteeing Non-repudiation. [Thesis]. Cochin University of Science and Technology; 2014. Available from: http://dyuthi.cusat.ac.in/purl/4959

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


Addis Ababa University

15. Bezawit, Wubishet. Noise Robust Speaker Verification using SVM based GMM Supervector .

Degree: 2013, Addis Ababa University

 This thesis shows research performed on the topic of text-independent speaker verification using support vector machine (SVM) based Gaussian mixture model (GMM) supervector classifier in… (more)

Subjects/Keywords: Text-Independent Speaker Verification; SVM; GMM supervector; GMM-UBM; multi-condition training; LibSVM

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Bezawit, W. (2013). Noise Robust Speaker Verification using SVM based GMM Supervector . (Thesis). Addis Ababa University. Retrieved from http://etd.aau.edu.et/dspace/handle/123456789/4707

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Bezawit, Wubishet. “Noise Robust Speaker Verification using SVM based GMM Supervector .” 2013. Thesis, Addis Ababa University. Accessed January 21, 2020. http://etd.aau.edu.et/dspace/handle/123456789/4707.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Bezawit, Wubishet. “Noise Robust Speaker Verification using SVM based GMM Supervector .” 2013. Web. 21 Jan 2020.

Vancouver:

Bezawit W. Noise Robust Speaker Verification using SVM based GMM Supervector . [Internet] [Thesis]. Addis Ababa University; 2013. [cited 2020 Jan 21]. Available from: http://etd.aau.edu.et/dspace/handle/123456789/4707.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Bezawit W. Noise Robust Speaker Verification using SVM based GMM Supervector . [Thesis]. Addis Ababa University; 2013. Available from: http://etd.aau.edu.et/dspace/handle/123456789/4707

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


Virginia Tech

16. Raman, Pujita. Speaker Identification and Verification Using Line Spectral Frequencies.

Degree: MS, Electrical and Computer Engineering, 2015, Virginia Tech

 State-of-the-art speaker identification and verification (SIV) systems provide near perfect performance under clean conditions. However, their performance deteriorates in the presence of background noise. Many… (more)

Subjects/Keywords: Speech; Speaker; Noise; Identification; Verification; Recognition; Feature; Line Spectral Frequency; Gaussian Mixture Model; Transition; Vowel

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Raman, P. (2015). Speaker Identification and Verification Using Line Spectral Frequencies. (Masters Thesis). Virginia Tech. Retrieved from http://hdl.handle.net/10919/52964

Chicago Manual of Style (16th Edition):

Raman, Pujita. “Speaker Identification and Verification Using Line Spectral Frequencies.” 2015. Masters Thesis, Virginia Tech. Accessed January 21, 2020. http://hdl.handle.net/10919/52964.

MLA Handbook (7th Edition):

Raman, Pujita. “Speaker Identification and Verification Using Line Spectral Frequencies.” 2015. Web. 21 Jan 2020.

Vancouver:

Raman P. Speaker Identification and Verification Using Line Spectral Frequencies. [Internet] [Masters thesis]. Virginia Tech; 2015. [cited 2020 Jan 21]. Available from: http://hdl.handle.net/10919/52964.

Council of Science Editors:

Raman P. Speaker Identification and Verification Using Line Spectral Frequencies. [Masters Thesis]. Virginia Tech; 2015. Available from: http://hdl.handle.net/10919/52964


University of New South Wales

17. Kua, Jia Min Karen. Improving automatic speaker verification using front-end and back-end diversity.

Degree: Electrical Engineering & Telecommunications, 2012, University of New South Wales

 Technologies that exploit biometrics can potentially be applied to the identification and verification of individuals for controlling access to secured areas or materials. Among these… (more)

Subjects/Keywords: Spectral centroid features; Automatic Speaker Verification; Group delay features; Normalised information distance; Sparse representation classification

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Kua, J. M. K. (2012). Improving automatic speaker verification using front-end and back-end diversity. (Doctoral Dissertation). University of New South Wales. Retrieved from http://handle.unsw.edu.au/1959.4/51906 ; https://unsworks.unsw.edu.au/fapi/datastream/unsworks:10576/SOURCE02?view=true

Chicago Manual of Style (16th Edition):

Kua, Jia Min Karen. “Improving automatic speaker verification using front-end and back-end diversity.” 2012. Doctoral Dissertation, University of New South Wales. Accessed January 21, 2020. http://handle.unsw.edu.au/1959.4/51906 ; https://unsworks.unsw.edu.au/fapi/datastream/unsworks:10576/SOURCE02?view=true.

MLA Handbook (7th Edition):

Kua, Jia Min Karen. “Improving automatic speaker verification using front-end and back-end diversity.” 2012. Web. 21 Jan 2020.

Vancouver:

Kua JMK. Improving automatic speaker verification using front-end and back-end diversity. [Internet] [Doctoral dissertation]. University of New South Wales; 2012. [cited 2020 Jan 21]. Available from: http://handle.unsw.edu.au/1959.4/51906 ; https://unsworks.unsw.edu.au/fapi/datastream/unsworks:10576/SOURCE02?view=true.

Council of Science Editors:

Kua JMK. Improving automatic speaker verification using front-end and back-end diversity. [Doctoral Dissertation]. University of New South Wales; 2012. Available from: http://handle.unsw.edu.au/1959.4/51906 ; https://unsworks.unsw.edu.au/fapi/datastream/unsworks:10576/SOURCE02?view=true

18. ΦΑΚΩΤΑΚΗΣ, ΝΙΚΟΛΑΟΣ. ΝΕΑ ΜΕΘΟΔΟΣ ΑΥΤΟΜΑΤΗΣ ΑΝΑΓΝΩΡΙΣΗΣ ΟΜΙΛΗΤΗ.

Degree: 1986, Πανεπιστήμιο Πατρών; University of Patras

ΣΤΗ ΔΙΑΤΡΙΒΗ ΠΑΡΟΥΣΙΑΖΕΤΑΙ ΜΙΑ ΝΕΑ ΜΕΘΟΔΟΣ ΑΥΤΟΜΑΤΗΣ ΑΝΑΓΝΩΡΙΣΕΩΣ ΟΜΙΛΗΤΗ (ΕΞΑΚΡΙΒΩΣΕΩΣ ΚΑΙ ΕΠΙΒΕΒΑΙΩΣΕΩΣ), ΠΟΥ ΒΑΣΙΖΕΤΑΙ ΣΕ ΕΙΔΙΚΑ ΕΠΙΛΕΓΜΕΝΟ ΚΕΙΜΕΝΟ ΟΜΙΛΙΑΣ. ΤΟ ΚΕΙΜΕΝΟ ΑΠΟΤΕΛΕΙΤΑΙ ΑΠΟ ΠΡΟΤΑΣΗ (#… (more)

Subjects/Keywords: AUTOMATIC SPEAKER RECOGNITION; SPEAKER IDENTIFICATION; Speaker recognition; Speaker verification; Αναγνώριση ομιλητή; ΑΥΤΟΜΑΤΗ ΜΕΘΟΔΟΣ ΑΝΑΓΝΩΡΙΣΕΩΣ ΟΜΙΛΗΤΗ; ΕΞΑΚΡΙΒΩΣΗ ΟΜΙΛΗΤΗ; Επιβεβαίωση ομιλητή

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

ΦΑΚΩΤΑΚΗΣ, . (1986). ΝΕΑ ΜΕΘΟΔΟΣ ΑΥΤΟΜΑΤΗΣ ΑΝΑΓΝΩΡΙΣΗΣ ΟΜΙΛΗΤΗ. (Thesis). Πανεπιστήμιο Πατρών; University of Patras. Retrieved from http://hdl.handle.net/10442/hedi/0048

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

ΦΑΚΩΤΑΚΗΣ, ΝΙΚΟΛΑΟΣ. “ΝΕΑ ΜΕΘΟΔΟΣ ΑΥΤΟΜΑΤΗΣ ΑΝΑΓΝΩΡΙΣΗΣ ΟΜΙΛΗΤΗ.” 1986. Thesis, Πανεπιστήμιο Πατρών; University of Patras. Accessed January 21, 2020. http://hdl.handle.net/10442/hedi/0048.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

ΦΑΚΩΤΑΚΗΣ, ΝΙΚΟΛΑΟΣ. “ΝΕΑ ΜΕΘΟΔΟΣ ΑΥΤΟΜΑΤΗΣ ΑΝΑΓΝΩΡΙΣΗΣ ΟΜΙΛΗΤΗ.” 1986. Web. 21 Jan 2020.

Vancouver:

ΦΑΚΩΤΑΚΗΣ . ΝΕΑ ΜΕΘΟΔΟΣ ΑΥΤΟΜΑΤΗΣ ΑΝΑΓΝΩΡΙΣΗΣ ΟΜΙΛΗΤΗ. [Internet] [Thesis]. Πανεπιστήμιο Πατρών; University of Patras; 1986. [cited 2020 Jan 21]. Available from: http://hdl.handle.net/10442/hedi/0048.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

ΦΑΚΩΤΑΚΗΣ . ΝΕΑ ΜΕΘΟΔΟΣ ΑΥΤΟΜΑΤΗΣ ΑΝΑΓΝΩΡΙΣΗΣ ΟΜΙΛΗΤΗ. [Thesis]. Πανεπιστήμιο Πατρών; University of Patras; 1986. Available from: http://hdl.handle.net/10442/hedi/0048

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

19. Σιαφαρίκας, Μιχαήλ. Αναγνώριση ομιλητή και ομιλίας με χρήση κυματιδίων.

Degree: 2010, University of Patras

Σκοπός της παρούσας διατριβής είναι η εκμετάλλευση των κυματιδίων με σκοπό την βελτίωση της απόδοσης συστημάτων αναγνώρισης ομιλητή και ομιλίας. Στα πλαίσια αυτά, εισάγονται τέσσερις… (more)

Subjects/Keywords: Αναγνώριση ομιλητή; Επιβεβαίωση ομιλητή; Αναγνώριση ομιλίας; Κυματίδια; Πακέτα κυματιδίων; Παράμετροι ομιλίας; Κρίσιμες ζώνες συχνοτήτων; 006.454; Speaker recognition; Speaker verification; Speech recognition; Wavelets; Wavelet packets; Speech features; Critical bands

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Σιαφαρίκας, . (2010). Αναγνώριση ομιλητή και ομιλίας με χρήση κυματιδίων. (Doctoral Dissertation). University of Patras. Retrieved from http://nemertes.lis.upatras.gr/jspui/handle/10889/3617

Chicago Manual of Style (16th Edition):

Σιαφαρίκας, Μιχαήλ. “Αναγνώριση ομιλητή και ομιλίας με χρήση κυματιδίων.” 2010. Doctoral Dissertation, University of Patras. Accessed January 21, 2020. http://nemertes.lis.upatras.gr/jspui/handle/10889/3617.

MLA Handbook (7th Edition):

Σιαφαρίκας, Μιχαήλ. “Αναγνώριση ομιλητή και ομιλίας με χρήση κυματιδίων.” 2010. Web. 21 Jan 2020.

Vancouver:

Σιαφαρίκας . Αναγνώριση ομιλητή και ομιλίας με χρήση κυματιδίων. [Internet] [Doctoral dissertation]. University of Patras; 2010. [cited 2020 Jan 21]. Available from: http://nemertes.lis.upatras.gr/jspui/handle/10889/3617.

Council of Science Editors:

Σιαφαρίκας . Αναγνώριση ομιλητή και ομιλίας με χρήση κυματιδίων. [Doctoral Dissertation]. University of Patras; 2010. Available from: http://nemertes.lis.upatras.gr/jspui/handle/10889/3617


Queensland University of Technology

20. Dean, David Brendan. Synchronous HMMs for audio-visual speech processing.

Degree: 2008, Queensland University of Technology

 Both human perceptual studies and automaticmachine-based experiments have shown that visual information from a speaker's mouth region can improve the robustness of automatic speech processing… (more)

Subjects/Keywords: speech processing; speech recognition; speaker recognition; speaker verification; multimodal; audio-visual; data fusion; pattern recognition; hidden Markov models; synchronous hidden Markov models

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Dean, D. B. (2008). Synchronous HMMs for audio-visual speech processing. (Thesis). Queensland University of Technology. Retrieved from https://eprints.qut.edu.au/17689/

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Dean, David Brendan. “Synchronous HMMs for audio-visual speech processing.” 2008. Thesis, Queensland University of Technology. Accessed January 21, 2020. https://eprints.qut.edu.au/17689/.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Dean, David Brendan. “Synchronous HMMs for audio-visual speech processing.” 2008. Web. 21 Jan 2020.

Vancouver:

Dean DB. Synchronous HMMs for audio-visual speech processing. [Internet] [Thesis]. Queensland University of Technology; 2008. [cited 2020 Jan 21]. Available from: https://eprints.qut.edu.au/17689/.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Dean DB. Synchronous HMMs for audio-visual speech processing. [Thesis]. Queensland University of Technology; 2008. Available from: https://eprints.qut.edu.au/17689/

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


Brno University of Technology

21. Görig, Jan. Multiplatformní aplikace pro verifikaci mluvčího .

Degree: 2010, Brno University of Technology

 Bakalářská práce se zabývá rozpoznáváním mluvčího bez znalosti textu sdělení. Zmiňuje dnes používané způsoby extrakce příznaků a jejich vyhodnocení pomocí směsice Gaussových hustotních funkcí. Praktickým… (more)

Subjects/Keywords: verifikace mluvčího; identifikace mluvčího; multiplatformní aplikace; PCM; MFCC; STG; GMM; UBM; BSAPI; Qt; speaker verification; speaker identification; multiplatform application; PCM; MFCC; STG; GMM; UBM; BSAPI; Qt

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Görig, J. (2010). Multiplatformní aplikace pro verifikaci mluvčího . (Thesis). Brno University of Technology. Retrieved from http://hdl.handle.net/11012/53003

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Görig, Jan. “Multiplatformní aplikace pro verifikaci mluvčího .” 2010. Thesis, Brno University of Technology. Accessed January 21, 2020. http://hdl.handle.net/11012/53003.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Görig, Jan. “Multiplatformní aplikace pro verifikaci mluvčího .” 2010. Web. 21 Jan 2020.

Vancouver:

Görig J. Multiplatformní aplikace pro verifikaci mluvčího . [Internet] [Thesis]. Brno University of Technology; 2010. [cited 2020 Jan 21]. Available from: http://hdl.handle.net/11012/53003.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Görig J. Multiplatformní aplikace pro verifikaci mluvčího . [Thesis]. Brno University of Technology; 2010. Available from: http://hdl.handle.net/11012/53003

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


Brno University of Technology

22. Fux, Jan. Textově závislé rozpoznávání mluvčího .

Degree: 2013, Brno University of Technology

 Cílem této bakalářské práce bylo navrhnout systém pro textově závislé rozpoznávání mluvčího. Bylo otestováno několik přístupů na databázi MIT, která obsahuje nahrávky průměrné délky 0,46s.… (more)

Subjects/Keywords: rozpoznávání mluvčího; DTW; foném; neuronová síť; GMM; rozpoznávání promluvy; fúze; DET křivka; EER; DCF; speaker recognition; speaker verification; DTW; phoneme; neural network; GMM; speech recognition; fusion; DET curve; EER; DCF

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Fux, J. (2013). Textově závislé rozpoznávání mluvčího . (Thesis). Brno University of Technology. Retrieved from http://hdl.handle.net/11012/55069

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Fux, Jan. “Textově závislé rozpoznávání mluvčího .” 2013. Thesis, Brno University of Technology. Accessed January 21, 2020. http://hdl.handle.net/11012/55069.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Fux, Jan. “Textově závislé rozpoznávání mluvčího .” 2013. Web. 21 Jan 2020.

Vancouver:

Fux J. Textově závislé rozpoznávání mluvčího . [Internet] [Thesis]. Brno University of Technology; 2013. [cited 2020 Jan 21]. Available from: http://hdl.handle.net/11012/55069.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Fux J. Textově závislé rozpoznávání mluvčího . [Thesis]. Brno University of Technology; 2013. Available from: http://hdl.handle.net/11012/55069

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


RMIT University

23. Hossan, M. Automatic speaker recognition dynamic feature identification and classification using distributed discrete cosine transform based mel frequency cepstral coefficients and fuzzy vector quantization.

Degree: 2011, RMIT University

 The Mel-Frequency Cepstral Coefficients (MFCC) feature extraction method is a leading approach for speech feature extraction and current research aims to identify performance enhancements. In… (more)

Subjects/Keywords: Fields of Research; Automatic Speech Recognition; Automatic Speaker Verification; Discrete Cosine Transform; Delta- Delta Mel Frequency Cepstrum Coefficient; Fuzzy Vector Quantization

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Hossan, M. (2011). Automatic speaker recognition dynamic feature identification and classification using distributed discrete cosine transform based mel frequency cepstral coefficients and fuzzy vector quantization. (Thesis). RMIT University. Retrieved from http://researchbank.rmit.edu.au/view/rmit:12308

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Hossan, M. “Automatic speaker recognition dynamic feature identification and classification using distributed discrete cosine transform based mel frequency cepstral coefficients and fuzzy vector quantization.” 2011. Thesis, RMIT University. Accessed January 21, 2020. http://researchbank.rmit.edu.au/view/rmit:12308.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Hossan, M. “Automatic speaker recognition dynamic feature identification and classification using distributed discrete cosine transform based mel frequency cepstral coefficients and fuzzy vector quantization.” 2011. Web. 21 Jan 2020.

Vancouver:

Hossan M. Automatic speaker recognition dynamic feature identification and classification using distributed discrete cosine transform based mel frequency cepstral coefficients and fuzzy vector quantization. [Internet] [Thesis]. RMIT University; 2011. [cited 2020 Jan 21]. Available from: http://researchbank.rmit.edu.au/view/rmit:12308.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Hossan M. Automatic speaker recognition dynamic feature identification and classification using distributed discrete cosine transform based mel frequency cepstral coefficients and fuzzy vector quantization. [Thesis]. RMIT University; 2011. Available from: http://researchbank.rmit.edu.au/view/rmit:12308

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


University of Southern California

24. George, Sageev Thycodam. The use of dynamic synapse neural networks for speech processing tasks.

Degree: PhD, Biomedical Engineering, 2007, University of Southern California

 Automated signal processing systems have become increasingly popular in the modern world. One challenge to developing such systems is the user's inability or lack of… (more)

Subjects/Keywords: neural network; speech recognition; speaker verification; dynamic synapse

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

George, S. T. (2007). The use of dynamic synapse neural networks for speech processing tasks. (Doctoral Dissertation). University of Southern California. Retrieved from http://digitallibrary.usc.edu/cdm/compoundobject/collection/p15799coll127/id/520664/rec/7378

Chicago Manual of Style (16th Edition):

George, Sageev Thycodam. “The use of dynamic synapse neural networks for speech processing tasks.” 2007. Doctoral Dissertation, University of Southern California. Accessed January 21, 2020. http://digitallibrary.usc.edu/cdm/compoundobject/collection/p15799coll127/id/520664/rec/7378.

MLA Handbook (7th Edition):

George, Sageev Thycodam. “The use of dynamic synapse neural networks for speech processing tasks.” 2007. Web. 21 Jan 2020.

Vancouver:

George ST. The use of dynamic synapse neural networks for speech processing tasks. [Internet] [Doctoral dissertation]. University of Southern California; 2007. [cited 2020 Jan 21]. Available from: http://digitallibrary.usc.edu/cdm/compoundobject/collection/p15799coll127/id/520664/rec/7378.

Council of Science Editors:

George ST. The use of dynamic synapse neural networks for speech processing tasks. [Doctoral Dissertation]. University of Southern California; 2007. Available from: http://digitallibrary.usc.edu/cdm/compoundobject/collection/p15799coll127/id/520664/rec/7378


Queensland University of Technology

25. Kanagasundaram, Ahilan. Speaker verification using I-vector features.

Degree: 2014, Queensland University of Technology

 This PhD research has provided novel solutions to three major challenges which have prevented the wide spread deployment of speaker recognition technology: (1) combating enrolment/… (more)

Subjects/Keywords: speaker verification; I-vectors; PLDA; LDA; WLDA; channel compensation; short utterance i-vectors; SN-WLDA; utterance varaince modelling; GPLDA

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Kanagasundaram, A. (2014). Speaker verification using I-vector features. (Thesis). Queensland University of Technology. Retrieved from https://eprints.qut.edu.au/77834/

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Kanagasundaram, Ahilan. “Speaker verification using I-vector features.” 2014. Thesis, Queensland University of Technology. Accessed January 21, 2020. https://eprints.qut.edu.au/77834/.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Kanagasundaram, Ahilan. “Speaker verification using I-vector features.” 2014. Web. 21 Jan 2020.

Vancouver:

Kanagasundaram A. Speaker verification using I-vector features. [Internet] [Thesis]. Queensland University of Technology; 2014. [cited 2020 Jan 21]. Available from: https://eprints.qut.edu.au/77834/.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Kanagasundaram A. Speaker verification using I-vector features. [Thesis]. Queensland University of Technology; 2014. Available from: https://eprints.qut.edu.au/77834/

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

26. Srinivasan, Sundararajan. A NONLINEAR MIXTURE AUTOREGRESSIVE MODEL FOR SPEAKER VERIFICATION.

Degree: PhD, Electrical and Computer Engineering, 2011, Mississippi State University

  In this work, we apply a nonlinear mixture autoregressive (MixAR) model to supplant the Gaussian mixture model for speaker verification. MixAR is a statistical… (more)

Subjects/Keywords: nonlinear model; speaker verification; mixture autoregressive model

…6.4 6.5 6.6 6.7 6.8 VII. Statistical Significance in Speaker Verification Systems… …82 Speaker Verification Experiments with TIMIT under Noisy Conditions… …84 Speaker Verification Experiments Across Channel Conditions – TIMIT vs. NTIMIT Database… …88 Speaker Verification Performance and Training Data Duration..........89 Speaker… …Verification Performance (EER) for different noise conditions ..........87 6.6 Speaker… 

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Srinivasan, S. (2011). A NONLINEAR MIXTURE AUTOREGRESSIVE MODEL FOR SPEAKER VERIFICATION. (Doctoral Dissertation). Mississippi State University. Retrieved from http://sun.library.msstate.edu/ETD-db/theses/available/etd-11302010-132852/ ;

Chicago Manual of Style (16th Edition):

Srinivasan, Sundararajan. “A NONLINEAR MIXTURE AUTOREGRESSIVE MODEL FOR SPEAKER VERIFICATION.” 2011. Doctoral Dissertation, Mississippi State University. Accessed January 21, 2020. http://sun.library.msstate.edu/ETD-db/theses/available/etd-11302010-132852/ ;.

MLA Handbook (7th Edition):

Srinivasan, Sundararajan. “A NONLINEAR MIXTURE AUTOREGRESSIVE MODEL FOR SPEAKER VERIFICATION.” 2011. Web. 21 Jan 2020.

Vancouver:

Srinivasan S. A NONLINEAR MIXTURE AUTOREGRESSIVE MODEL FOR SPEAKER VERIFICATION. [Internet] [Doctoral dissertation]. Mississippi State University; 2011. [cited 2020 Jan 21]. Available from: http://sun.library.msstate.edu/ETD-db/theses/available/etd-11302010-132852/ ;.

Council of Science Editors:

Srinivasan S. A NONLINEAR MIXTURE AUTOREGRESSIVE MODEL FOR SPEAKER VERIFICATION. [Doctoral Dissertation]. Mississippi State University; 2011. Available from: http://sun.library.msstate.edu/ETD-db/theses/available/etd-11302010-132852/ ;

27. Wan, Qianhui. Speaker Verification Systems Under Various Noise and SNR Conditions .

Degree: 2017, University of Ottawa

 In speaker verification, the mismatches between the training speech and the testing speech can greatly affect the robustness of classification algorithms, and the mismatches are… (more)

Subjects/Keywords: Speaker verification; Various noise; SNR conditions

…It consists of two main topics: speaker identification and speaker verification. Given a… …speaker that sounds closest to the new speech. The task of speaker verification is to decide if… …and this task can be accomplished by speaker verification. Besides the application on… …identity authentication, speaker verification also has a vital role in forensics and law… …recognized, and speaker verification techniques are able to make a reliable decision from the… 

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Wan, Q. (2017). Speaker Verification Systems Under Various Noise and SNR Conditions . (Thesis). University of Ottawa. Retrieved from http://hdl.handle.net/10393/36888

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Wan, Qianhui. “Speaker Verification Systems Under Various Noise and SNR Conditions .” 2017. Thesis, University of Ottawa. Accessed January 21, 2020. http://hdl.handle.net/10393/36888.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Wan, Qianhui. “Speaker Verification Systems Under Various Noise and SNR Conditions .” 2017. Web. 21 Jan 2020.

Vancouver:

Wan Q. Speaker Verification Systems Under Various Noise and SNR Conditions . [Internet] [Thesis]. University of Ottawa; 2017. [cited 2020 Jan 21]. Available from: http://hdl.handle.net/10393/36888.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Wan Q. Speaker Verification Systems Under Various Noise and SNR Conditions . [Thesis]. University of Ottawa; 2017. Available from: http://hdl.handle.net/10393/36888

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

28. PINHEIRO, Hector Natan Batista. Verificação de locutores independente de texto: uma análise de robustez a ruído .

Degree: 2015, Universidade Federal de Pernambuco

 O processo de identificação de um determinado indivíduo é realizado milhões de vezes, todos os dias, por organizações dos mais diversos setores. Perguntas como "Quem… (more)

Subjects/Keywords: Identificação biométrica. Reconhecimento de locutores. Sistemas de verificação de locutores independente de texto. Robustez de sistemas de reconhecimento de locutores. Técnicas de compensação.; Speaker recognition. Text-independent speaker verification systems. Robust speaker recognition systems. Compensation techniques

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

PINHEIRO, H. N. B. (2015). Verificação de locutores independente de texto: uma análise de robustez a ruído . (Masters Thesis). Universidade Federal de Pernambuco. Retrieved from https://repositorio.ufpe.br/handle/123456789/18045

Chicago Manual of Style (16th Edition):

PINHEIRO, Hector Natan Batista. “Verificação de locutores independente de texto: uma análise de robustez a ruído .” 2015. Masters Thesis, Universidade Federal de Pernambuco. Accessed January 21, 2020. https://repositorio.ufpe.br/handle/123456789/18045.

MLA Handbook (7th Edition):

PINHEIRO, Hector Natan Batista. “Verificação de locutores independente de texto: uma análise de robustez a ruído .” 2015. Web. 21 Jan 2020.

Vancouver:

PINHEIRO HNB. Verificação de locutores independente de texto: uma análise de robustez a ruído . [Internet] [Masters thesis]. Universidade Federal de Pernambuco; 2015. [cited 2020 Jan 21]. Available from: https://repositorio.ufpe.br/handle/123456789/18045.

Council of Science Editors:

PINHEIRO HNB. Verificação de locutores independente de texto: uma análise de robustez a ruído . [Masters Thesis]. Universidade Federal de Pernambuco; 2015. Available from: https://repositorio.ufpe.br/handle/123456789/18045


Indian Institute of Science

29. Abhiram, B. Characterization of the Voice Source by the DCT for Speaker Information.

Degree: 2014, Indian Institute of Science

 Extracting speaker-specific information from speech is of great interest to both researchers and developers alike, since speaker recognition technology finds application in a wide range… (more)

Subjects/Keywords: Speaker Recognition; Voice Source Characterization; Speech Source Filter Model; Speech Processing; Discrete Cosine Transform (DCT); Speaker Information; Integrated Linear Prediction Residual (ILPR); Digital Signal Processing; Speech Signals; Speech Recognition; Speech Production; Source-filter Model; Speaker Verification; Electrical Engineering

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Abhiram, B. (2014). Characterization of the Voice Source by the DCT for Speaker Information. (Thesis). Indian Institute of Science. Retrieved from http://etd.iisc.ernet.in/handle/2005/2894 ; http://etd.ncsi.iisc.ernet.in/abstracts/3756/G26314-Abs.pdf

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Abhiram, B. “Characterization of the Voice Source by the DCT for Speaker Information.” 2014. Thesis, Indian Institute of Science. Accessed January 21, 2020. http://etd.iisc.ernet.in/handle/2005/2894 ; http://etd.ncsi.iisc.ernet.in/abstracts/3756/G26314-Abs.pdf.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Abhiram, B. “Characterization of the Voice Source by the DCT for Speaker Information.” 2014. Web. 21 Jan 2020.

Vancouver:

Abhiram B. Characterization of the Voice Source by the DCT for Speaker Information. [Internet] [Thesis]. Indian Institute of Science; 2014. [cited 2020 Jan 21]. Available from: http://etd.iisc.ernet.in/handle/2005/2894 ; http://etd.ncsi.iisc.ernet.in/abstracts/3756/G26314-Abs.pdf.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Abhiram B. Characterization of the Voice Source by the DCT for Speaker Information. [Thesis]. Indian Institute of Science; 2014. Available from: http://etd.iisc.ernet.in/handle/2005/2894 ; http://etd.ncsi.iisc.ernet.in/abstracts/3756/G26314-Abs.pdf

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation


Indian Institute of Science

30. Abhiram, B. Characterization of the Voice Source by the DCT for Speaker Information.

Degree: 2014, Indian Institute of Science

 Extracting speaker-specific information from speech is of great interest to both researchers and developers alike, since speaker recognition technology finds application in a wide range… (more)

Subjects/Keywords: Speaker Recognition; Voice Source Characterization; Speech Source Filter Model; Speech Processing; Discrete Cosine Transform (DCT); Speaker Information; Integrated Linear Prediction Residual (ILPR); Digital Signal Processing; Speech Signals; Speech Recognition; Speech Production; Source-filter Model; Speaker Verification; Electrical Engineering

Record DetailsSimilar RecordsGoogle PlusoneFacebookTwitterCiteULikeMendeleyreddit

APA · Chicago · MLA · Vancouver · CSE | Export to Zotero / EndNote / Reference Manager

APA (6th Edition):

Abhiram, B. (2014). Characterization of the Voice Source by the DCT for Speaker Information. (Thesis). Indian Institute of Science. Retrieved from http://hdl.handle.net/2005/2894

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Chicago Manual of Style (16th Edition):

Abhiram, B. “Characterization of the Voice Source by the DCT for Speaker Information.” 2014. Thesis, Indian Institute of Science. Accessed January 21, 2020. http://hdl.handle.net/2005/2894.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

MLA Handbook (7th Edition):

Abhiram, B. “Characterization of the Voice Source by the DCT for Speaker Information.” 2014. Web. 21 Jan 2020.

Vancouver:

Abhiram B. Characterization of the Voice Source by the DCT for Speaker Information. [Internet] [Thesis]. Indian Institute of Science; 2014. [cited 2020 Jan 21]. Available from: http://hdl.handle.net/2005/2894.

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

Council of Science Editors:

Abhiram B. Characterization of the Voice Source by the DCT for Speaker Information. [Thesis]. Indian Institute of Science; 2014. Available from: http://hdl.handle.net/2005/2894

Note: this citation may be lacking information needed for this citation format:
Not specified: Masters Thesis or Doctoral Dissertation

[1] [2]

.