ROBUST HYBRID FEATURES BASED TEXT INDEPENDENT SPEAKER IDENTIFICATION SYSTEM OVER NOISY ADDITIVE CHANNEL
Robustness of speaker identification systems over additive noise is crucial for real-world applications. In this paper, two robust features named Power Normalized Cepstral Coefficients (PNCC) and Gammatone Frequency Cepstral Coefficients (GFCC) are combined together to improve the robustness of spe...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | Arabic |
| Published: |
Mustansiriyah University/College of Engineering
2020-07-01
|
| Series: | Journal of Engineering and Sustainable Development |
| Subjects: | |
| Online Access: | https://jeasd.uomustansiriyah.edu.iq/index.php/jeasd/article/view/117 |
| _version_ | 1797683295664734208 |
|---|---|
| author | Ali Muayad Jalil Fadhel Sahib Hasan Hesham Adnan Alabbasi |
| author_facet | Ali Muayad Jalil Fadhel Sahib Hasan Hesham Adnan Alabbasi |
| author_sort | Ali Muayad Jalil |
| collection | DOAJ |
| description |
Robustness of speaker identification systems over additive noise is crucial for real-world applications. In this paper, two robust features named Power Normalized Cepstral Coefficients (PNCC) and Gammatone Frequency Cepstral Coefficients (GFCC) are combined together to improve the robustness of speaker identification system over different types of noise. Universal Background Model Gaussian Mixture Model (UBM-GMM) is used as a feature matching and a classifier to identify the claim speakers. Evaluation results show that the proposed hybrid feature improves the performance of identification system when compared to conventional features over most types of noise and different signal-to-noise ratios.
|
| first_indexed | 2024-03-12T00:12:27Z |
| format | Article |
| id | doaj.art-c87cd544077a442ea6b76621027769aa |
| institution | Directory Open Access Journal |
| issn | 2520-0917 2520-0925 |
| language | Arabic |
| last_indexed | 2024-03-12T00:12:27Z |
| publishDate | 2020-07-01 |
| publisher | Mustansiriyah University/College of Engineering |
| record_format | Article |
| series | Journal of Engineering and Sustainable Development |
| spelling | doaj.art-c87cd544077a442ea6b76621027769aa2023-09-15T22:01:15ZaraMustansiriyah University/College of EngineeringJournal of Engineering and Sustainable Development2520-09172520-09252020-07-01244ROBUST HYBRID FEATURES BASED TEXT INDEPENDENT SPEAKER IDENTIFICATION SYSTEM OVER NOISY ADDITIVE CHANNELAli Muayad Jalil0Fadhel Sahib Hasan1Hesham Adnan Alabbasi2Electrical Engineering Department, Mustansiriyah University, Baghdad, IraqElectrical Engineering Department, Mustansiriyah University, Baghdad, IraqComputer Engineering Department, Mustansiriyah University, Baghdad, Iraq Robustness of speaker identification systems over additive noise is crucial for real-world applications. In this paper, two robust features named Power Normalized Cepstral Coefficients (PNCC) and Gammatone Frequency Cepstral Coefficients (GFCC) are combined together to improve the robustness of speaker identification system over different types of noise. Universal Background Model Gaussian Mixture Model (UBM-GMM) is used as a feature matching and a classifier to identify the claim speakers. Evaluation results show that the proposed hybrid feature improves the performance of identification system when compared to conventional features over most types of noise and different signal-to-noise ratios. https://jeasd.uomustansiriyah.edu.iq/index.php/jeasd/article/view/117robust speaker identificationrobust feature extractionPNCCGFCCFWUBM-GMM |
| spellingShingle | Ali Muayad Jalil Fadhel Sahib Hasan Hesham Adnan Alabbasi ROBUST HYBRID FEATURES BASED TEXT INDEPENDENT SPEAKER IDENTIFICATION SYSTEM OVER NOISY ADDITIVE CHANNEL Journal of Engineering and Sustainable Development robust speaker identification robust feature extraction PNCC GFCC FW UBM-GMM |
| title | ROBUST HYBRID FEATURES BASED TEXT INDEPENDENT SPEAKER IDENTIFICATION SYSTEM OVER NOISY ADDITIVE CHANNEL |
| title_full | ROBUST HYBRID FEATURES BASED TEXT INDEPENDENT SPEAKER IDENTIFICATION SYSTEM OVER NOISY ADDITIVE CHANNEL |
| title_fullStr | ROBUST HYBRID FEATURES BASED TEXT INDEPENDENT SPEAKER IDENTIFICATION SYSTEM OVER NOISY ADDITIVE CHANNEL |
| title_full_unstemmed | ROBUST HYBRID FEATURES BASED TEXT INDEPENDENT SPEAKER IDENTIFICATION SYSTEM OVER NOISY ADDITIVE CHANNEL |
| title_short | ROBUST HYBRID FEATURES BASED TEXT INDEPENDENT SPEAKER IDENTIFICATION SYSTEM OVER NOISY ADDITIVE CHANNEL |
| title_sort | robust hybrid features based text independent speaker identification system over noisy additive channel |
| topic | robust speaker identification robust feature extraction PNCC GFCC FW UBM-GMM |
| url | https://jeasd.uomustansiriyah.edu.iq/index.php/jeasd/article/view/117 |
| work_keys_str_mv | AT alimuayadjalil robusthybridfeaturesbasedtextindependentspeakeridentificationsystemovernoisyadditivechannel AT fadhelsahibhasan robusthybridfeaturesbasedtextindependentspeakeridentificationsystemovernoisyadditivechannel AT heshamadnanalabbasi robusthybridfeaturesbasedtextindependentspeakeridentificationsystemovernoisyadditivechannel |