Robust speaker verification by combining MFCC and entrocy in noisy conditions

Mohammed, Duraid Y.; Al-Karawi, Khamis A.; Aljuboori, Ahmed

doi:10.11591/eei.v10i4.2957

Cited by 10 publications

(6 citation statements)

References 21 publications

(30 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To detect more differences between genuine and spoof voices in the voice-transformed images such as spectrogram, image classification techniques could be used as suggested in [17]. MFCC is an audio feature commonly used for signal processing, especially speech recognition [18], [19]. Figure 5 shows the generated MFCC images of genuine and spoof voices.…”

Section: The Proposed Methodsmentioning

confidence: 99%

Artificial speech detection using image-based features and random forest classifier

Tan

Hijazi

Kok³

et al. 2022

IJ-AI

View full text Add to dashboard Cite

The ASVspoof 2015 Challenge was one of the efforts of the research community in the field of speech processing to foster the development of generalized countermeasures against spoofing attacks. However, most countermeasures submitted to the ASVspoof 2015 Challenge failed to detect the S10 attack effectively, the only attack that was generated using the waveform concatenation approach. Hence, more informative features are needed to detect previously unseen spoofing attacks. This paper presents an approach that uses data transformation techniques to engineer image-based features together with random forest classifier to detect artificial speech. The objectives are two-fold: (i) to extract image-based features from the melfrequency cepstral coefficients representation of the speech signal and (ii) to compare the performance of using the extracted features and Random Forest to determine the authenticity of voices with the existing approaches. An audio-to-image transformation technique was used to engineer new features in classifying genuine and spoof voices. An experiment was conducted to find the appropriate combination of the engineered features and classifier. Experimental results showed that the proposed approach was able to detect speech synthesis and voice conversion attacks effectively, with an equal error rate of 0.10% and accuracy of 99.93%.

show abstract

Section: The Proposed Methodsmentioning

confidence: 99%

Artificial speech detection using image-based features and random forest classifier

Tan

Hijazi

Kok³

et al. 2022

IJ-AI

View full text Add to dashboard Cite

show abstract

“…Figure (10) These parameters are also shown to be successful, across a wide variety of classification systems. The process through which the MFCC is calculated can be described as follows [13], [18]:…”

Section: -Spectral Rolloffmentioning

confidence: 99%

Real-Time Lie-Speech Determination Using Voice-Stress Technology

Al-Dhaher,

Y. Mohammed,

Khalaf

et al. 2024

Iraqi Journal For Computer Science and Mathematics

View full text Add to dashboard Cite

Lie detection has gained importance and is now extremely significant in a variety of fields. It playsan important role in several domains, including law enforcement, criminal investigations, national security,workplace ethics, and personal relationships. As advances in lie detection continue to develop, real-timeapproaches such as voice stress technology have emerged as a feasible alternative to traditional methods such aspolygraph testing. Polygraph testing, a historical and generally established approach, may be enhanced or replacedby these revolutionary real-time techniques. Traditional lie detection procedures, such as polygraph testing, havebeen challenged for their lack of reliability and validity. Newer techniques, such as brain imaging and machinelearning, might offer better outcomes, although they are still in their early phases and require additional testing.This project intends to explore a deception-detection module based on sophisticated speech-stress analysistechniques that might be applied in a real-time deception system. The purpose is to study stress and otherarticulation cues in voice patterns, to establish their precision and reliability in detecting deceit, by building uponprevious knowledge and applying state-of-the-art architecture. The performance and accuracy of the system and itsaudio aspects will be thoroughly analyzed. The ultimate purpose is to contribute to the advancement of moreaccurate and reliable lie-detection systems, by addressing the limitations of old techniques and proposing practicalsolutions for varied applications. This paper proposes an efficient feature-selection strategy, which uses randomforest (RF) to select only the significant features for training when a real-life trial dataset consisting of audio filesis employed. Next, utilizing the RF as a classifier, an accuracy of 88% is reached through comprehensiveevaluation, thereby confirming its reliability and precision for lie-detection in real-time scenarios.

show abstract

“…Several studies proved that integrating algorithms can increase performance compared to original algorithms [17][18][19] and 20 . This paper is an attempt to combine two cryptography algorithms.…”

Section: Brief Of Related Workmentioning

confidence: 99%

Digital Data Encryption Using a Proposed W-Method Based on AES and DES Algorithms

Shukur

Qurban²,

Aljuboori³

2023

Baghdad Sci.J

View full text Add to dashboard Cite

This paper proposes a new encryption method. It combines two cipher algorithms, i.e., DES and AES, to generate hybrid keys. This combination strengthens the proposed W-method by generating high randomized keys. Two points can represent the reliability of any encryption technique. Firstly, is the key generation; therefore, our approach merges 64 bits of DES with 64 bits of AES to produce 128 bits as a root key for all remaining keys that are 15. This complexity increases the level of the ciphering process. Moreover, it shifts the operation one bit only to the right. Secondly is the nature of the encryption process. It includes two keys and mixes one round of DES with one round of AES to reduce the performance time. The W-method deals with Arabic and English texts with the same efficiency. The result showed that the proposed method performs faster and more securely when compared to standard DES and AES algorithms.

show abstract

Robust speaker verification by combining MFCC and entrocy in noisy conditions

Cited by 10 publications

References 21 publications

Artificial speech detection using image-based features and random forest classifier

Artificial speech detection using image-based features and random forest classifier

Real-Time Lie-Speech Determination Using Voice-Stress Technology

Digital Data Encryption Using a Proposed W-Method Based on AES and DES Algorithms

Contact Info

Product

Resources

About