Search

In the realm of speaker identification, pitch frequency serves as a fundamental feature. However, this feature can be compromised when a speaker records his speech in a closed room, resulting in distorted signal features. This distortion not only reduces the effectiveness of speaker identification systems, but also opens the door for potential deception by hackers who exploit the reverberation effects in closed rooms. To address this concern, the correction of estimated pitch frequencies emerges as an essential step for the success of speaker identification systems. This paper presents a Hybrid Approach for Estimating Pitch Frequency (HAEPF) that integrates both the Zero Crossing Rate (ZCR) and Auto-Correlation Function (ACF) methods. Furthermore, the paper delves into the modeling of reverberant speech using comb filtering, shedding light on how multiple reflections impact the accuracy of pitch frequency estimation. Several simulation experiments were conducted to assess pitch frequency estimation for speech signals, both in the presence and absence of reverberation. The estimation errors were calculated for all three scenarios of reverberation (mild, moderate, and severe). The results clearly indicate that as the degree of reverberation, characterized by the comb filter order, increases, the pitch frequency estimation error also increases. The estimation accuracy of the proposed approach is calculated in terms of Pitch Frequency Estimation Error (PFEE), Gross Pitch Error (GPE) and Octave Error (OER) and is compared with those of several established pitch frequency estimation methods. The proposed approach exhibits a notable enhancement even in noisy environments, reducing PFEE by 43%, and achieving GPE and OER of less than 0.3 and 0.12, respectively, at a Signal-to-Noise Ratio (SNR) of 0 dB. [ABSTRACT FROM AUTHOR]

This paper introduces a method aiming at enhancing the efficacy of speaker identification systems within challenging acoustic environments characterized by noise and reverberation. The methodology encompasses the utilization of diverse feature extraction techniques, including Mel-Frequency Cepstral Coefficients (MFCCs) and discrete transforms, such as Discrete Cosine Transform (DCT), Discrete Sine Transform (DST), and Discrete Wavelet Transform (DWT). Additionally, an Artificial Neural Network (ANN) serves as the classifier for this method. Reverberation is modeled using varying-length comb filters, and its impact on pitch frequency estimation is explored via the Auto Correlation Function (ACF). This paper also contributes to the field of cancelable speaker identification in both open and reverberation environments. The proposed method depends on comb filtering at the feature level, deliberately distorting MFCCs. This distortion, incorporated within a cancelable framework, serves to obscure speaker identities, rendering the system resilient to potential intruders. Three systems are presented in this work; a reverberation-affected speaker identification system, a system depending on cancelable features through comb filtering, and a novel cancelable speaker identification system within reverbration environments. The findings revealed that, in both scenarios with and without reverberation effects, the DWT-based features exhibited superior performance within the speaker identification system. Conversely, within the cancelable speaker identification system, the DCT-based features represent the top-performing choice. [ABSTRACT FROM AUTHOR]

Speaker recognition is one of several biometric recognition systems owing to its high importance in numerous applications of security and telecommunications. The key aspiration of speaker recognition systems is to know who is speaking depending on voice characteristics. This paper presents an extensive study of speaker recognition in both text-dependent and text-independent cases. Convolutional Neural Network (CNN) based feature extraction is extended to the text-dependent and text-independent speaker recognition tasks. In addition, the effect of reverberation on the speaker recognition system is addressed. All speech signals are converted into images by obtaining their spectrograms. Two proposed CNN models are presented for efficient speaker recognition from clean and reverberant speech signals. They depend on image processing concepts applied on spectrograms of speech signals. One of the proposed models is compared with a conventional Benchmark model in the text-independent scenario. The performance of the recognition system is measured by the recognition rate in the cases of clean and reverberant speech. [ABSTRACT FROM AUTHOR]