Gaussian-Filtered High-Frequency-Feature Trained Optimized BiLSTM Network for Spoofed-Speech Classification
Voice-controlled devices are in demand due to their hands-free controls. However, using voice-controlled devices in sensitive scenarios like smartphone applications and financial transactions requires protection against fraudulent attacks referred to as “speech spoofing”. The algorithms used in spoo...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-07-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/23/14/6637 |
| _version_ | 1797587463892369408 |
|---|---|
| author | Hiren Mewada Jawad F. Al-Asad Faris A. Almalki Adil H. Khan Nouf Abdullah Almujally Samir El-Nakla Qamar Naith |
| author_facet | Hiren Mewada Jawad F. Al-Asad Faris A. Almalki Adil H. Khan Nouf Abdullah Almujally Samir El-Nakla Qamar Naith |
| author_sort | Hiren Mewada |
| collection | DOAJ |
| description | Voice-controlled devices are in demand due to their hands-free controls. However, using voice-controlled devices in sensitive scenarios like smartphone applications and financial transactions requires protection against fraudulent attacks referred to as “speech spoofing”. The algorithms used in spoof attacks are practically unknown; hence, further analysis and development of spoof-detection models for improving spoof classification are required. A study of the spoofed-speech spectrum suggests that high-frequency features are able to discriminate genuine speech from spoofed speech well. Typically, linear or triangular filter banks are used to obtain high-frequency features. However, a Gaussian filter can extract more global information than a triangular filter. In addition, MFCC features are preferable among other speech features because of their lower covariance. Therefore, in this study, the use of a Gaussian filter is proposed for the extraction of inverted MFCC (iMFCC) features, providing high-frequency features. Complementary features are integrated with iMFCC to strengthen the features that aid in the discrimination of spoof speech. Deep learning has been proven to be efficient in classification applications, but the selection of its hyper-parameters and architecture is crucial and directly affects performance. Therefore, a Bayesian algorithm is used to optimize the BiLSTM network. Thus, in this study, we build a high-frequency-based optimized BiLSTM network to classify the spoofed-speech signal, and we present an extensive investigation using the ASVSpoof 2017 dataset. The optimized BiLSTM model is successfully trained with the least epoch and achieved a 99.58% validation accuracy. The proposed algorithm achieved a 6.58% EER on the evaluation dataset, with a relative improvement of 78% on a baseline spoof-identification system. |
| first_indexed | 2024-03-11T00:39:20Z |
| format | Article |
| id | doaj.art-a148459aad2244708927795fea252a9b |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-11T00:39:20Z |
| publishDate | 2023-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-a148459aad2244708927795fea252a9b2023-11-18T21:20:31ZengMDPI AGSensors1424-82202023-07-012314663710.3390/s23146637Gaussian-Filtered High-Frequency-Feature Trained Optimized BiLSTM Network for Spoofed-Speech ClassificationHiren Mewada0Jawad F. Al-Asad1Faris A. Almalki2Adil H. Khan3Nouf Abdullah Almujally4Samir El-Nakla5Qamar Naith6Electrical Engineering Department, Prince Mohammad bin Fahd University, P.O. Box 1664, Al Khobar 31952, Saudi ArabiaElectrical Engineering Department, Prince Mohammad bin Fahd University, P.O. Box 1664, Al Khobar 31952, Saudi ArabiaDepartment of Computer Engineering, College of Computers and Information Technology, Taif University, P.O. Box 11099, Taif 21944, Saudi ArabiaElectrical Engineering Department, Prince Mohammad bin Fahd University, P.O. Box 1664, Al Khobar 31952, Saudi ArabiaDepartment of Information Systems, College of Computer and Information Sciences, Princess Nourah bint Abdulrahman University, P.O. Box 84428, Riyadh 11671, Saudi ArabiaElectrical Engineering Department, Prince Mohammad bin Fahd University, P.O. Box 1664, Al Khobar 31952, Saudi ArabiaDepartment of Software Engineering, College of Computer Science and Engineering, University of Jeddah, P.O. Box 34, Jeddah 21959, Saudi ArabiaVoice-controlled devices are in demand due to their hands-free controls. However, using voice-controlled devices in sensitive scenarios like smartphone applications and financial transactions requires protection against fraudulent attacks referred to as “speech spoofing”. The algorithms used in spoof attacks are practically unknown; hence, further analysis and development of spoof-detection models for improving spoof classification are required. A study of the spoofed-speech spectrum suggests that high-frequency features are able to discriminate genuine speech from spoofed speech well. Typically, linear or triangular filter banks are used to obtain high-frequency features. However, a Gaussian filter can extract more global information than a triangular filter. In addition, MFCC features are preferable among other speech features because of their lower covariance. Therefore, in this study, the use of a Gaussian filter is proposed for the extraction of inverted MFCC (iMFCC) features, providing high-frequency features. Complementary features are integrated with iMFCC to strengthen the features that aid in the discrimination of spoof speech. Deep learning has been proven to be efficient in classification applications, but the selection of its hyper-parameters and architecture is crucial and directly affects performance. Therefore, a Bayesian algorithm is used to optimize the BiLSTM network. Thus, in this study, we build a high-frequency-based optimized BiLSTM network to classify the spoofed-speech signal, and we present an extensive investigation using the ASVSpoof 2017 dataset. The optimized BiLSTM model is successfully trained with the least epoch and achieved a 99.58% validation accuracy. The proposed algorithm achieved a 6.58% EER on the evaluation dataset, with a relative improvement of 78% on a baseline spoof-identification system.https://www.mdpi.com/1424-8220/23/14/6637anti-spoofingASVspoofconvolutional neural networkgenuine speech detectionvoice conversion |
| spellingShingle | Hiren Mewada Jawad F. Al-Asad Faris A. Almalki Adil H. Khan Nouf Abdullah Almujally Samir El-Nakla Qamar Naith Gaussian-Filtered High-Frequency-Feature Trained Optimized BiLSTM Network for Spoofed-Speech Classification Sensors anti-spoofing ASVspoof convolutional neural network genuine speech detection voice conversion |
| title | Gaussian-Filtered High-Frequency-Feature Trained Optimized BiLSTM Network for Spoofed-Speech Classification |
| title_full | Gaussian-Filtered High-Frequency-Feature Trained Optimized BiLSTM Network for Spoofed-Speech Classification |
| title_fullStr | Gaussian-Filtered High-Frequency-Feature Trained Optimized BiLSTM Network for Spoofed-Speech Classification |
| title_full_unstemmed | Gaussian-Filtered High-Frequency-Feature Trained Optimized BiLSTM Network for Spoofed-Speech Classification |
| title_short | Gaussian-Filtered High-Frequency-Feature Trained Optimized BiLSTM Network for Spoofed-Speech Classification |
| title_sort | gaussian filtered high frequency feature trained optimized bilstm network for spoofed speech classification |
| topic | anti-spoofing ASVspoof convolutional neural network genuine speech detection voice conversion |
| url | https://www.mdpi.com/1424-8220/23/14/6637 |
| work_keys_str_mv | AT hirenmewada gaussianfilteredhighfrequencyfeaturetrainedoptimizedbilstmnetworkforspoofedspeechclassification AT jawadfalasad gaussianfilteredhighfrequencyfeaturetrainedoptimizedbilstmnetworkforspoofedspeechclassification AT farisaalmalki gaussianfilteredhighfrequencyfeaturetrainedoptimizedbilstmnetworkforspoofedspeechclassification AT adilhkhan gaussianfilteredhighfrequencyfeaturetrainedoptimizedbilstmnetworkforspoofedspeechclassification AT noufabdullahalmujally gaussianfilteredhighfrequencyfeaturetrainedoptimizedbilstmnetworkforspoofedspeechclassification AT samirelnakla gaussianfilteredhighfrequencyfeaturetrainedoptimizedbilstmnetworkforspoofedspeechclassification AT qamarnaith gaussianfilteredhighfrequencyfeaturetrainedoptimizedbilstmnetworkforspoofedspeechclassification |