Speech Spectrum Research Articles

An Analysis of Emotional Responses of Students in Bilingual Classes and Adjustment Strategies

Students’ willingness to participate in bilingual communication is greatly influenced by their positive emotions in the bilingual class. The automatic recognition of students’ emotional state in bilingual class can assist teachers to correctly master the laws of students’ emotional changes during the bilingual learning process as fast as possible. However, the speech emotion features extracted by existing speech emotion recognition models are not universal and not suitable for bilingual speech emotion recognition, and the accuracy needs to be improved. To cope with these issues, this paper aims to study the emotional responses of students in bilingual class based on emotional analysis and provide adjustment strategies for it. At first, the signals of students’ speech records in bilingual class were pre-processed, the one-dimensional features of the signals were converted into 2-dimensional speech spectrum features so as to attain more useful information to facilitate the emotional recognition of students’ speech records in the bilingual class. Then this paper combined the bi-linear Convolution Neural Network (CNN) with capsule network to construct a bilingual class student emotion recognition model, and experimental results verified the effectiveness of the constructed model.

SOCIALIZATION OF A CHILD WITH SPECIAL NEEDS

In the modern world the problem of children inclusion has become very acute. Possibility to communicate is in high demand among children with special needs. Speech barriers can easily be overcome by using social networks and various chat rooms for communication. With virtual communication, shyness and avoidance of communication are reduced for both normal children and children with special needs. The objective of this study is to reveal positive tendency in communication on the Internet among children with special needs. The method of the study is the questionnaire which has been elaborated and adapted for the respondents. The respondents are  teenagers with special needs aged from  to . Main results show that  respondents out of  have not changed their views on communication with their peers. Other  respondents have felt that shyness in communication has been reduced. The remaining  people have had positive interaction experience. They are willing to continue communication in social networks and have found online pen pals. Conclusion. The study has showed that social networking and messaging have a positive impact on the attitude of adolescents with special needs towards communication. Many children with special needs have speech spectrum disorders and have diffi culties in starting informal communication with peers because of shyness and increased time of this communication. However, in virtual communication, these barriers are signifi cantly reduced, the quality of social life, confi dence, overall psychological state are improved. In addition, it is important that respondents “on the other side of the screen” (normal teenagers) have been willing to make contact, showed interest (which might not have happened in everyday life) even if they never had prior communication with people with special needs. Therefore, virtual communication is effective as a way of establishing informal communication for children with special needs.

Comparison of Two Place-Based Mapping Procedures on Masked Sentence Recognition as a Function of Electrode Array Angular Insertion Depth and Presence of Acoustic Low-Frequency Information: A Simulation Study

Introduction: Cochlear implant (CI) and electric-acoustic stimulation (EAS) users may experience better performance with maps that align the electric filter frequencies to the cochlear place frequencies, known as place-based maps, than with maps that present spectrally shifted information. Individual place-based mapping procedures differ in the frequency content that is aligned to cochlear tonotopicity versus discarded or spectrally shifted. The performance benefit with different place-based maps may vary due to individual differences in angular insertion depth (AID) of the electrode array and whether functional acoustic low-frequency information is available in the implanted ear. The present study compared masked speech recognition with two types of place-based maps as a function of AID and presence of acoustic low-frequency information. Methods: Sixty adults with normal hearing listened acutely to CI or EAS simulations of two types of place-based maps for one of three cases of electrode arrays at shallow AIDs. The strict place-based (Strict-PB) map aligned the low- and mid-frequency information to cochlear tonotopicity and discarded information below the frequency associated with the most apical electrode contact. The alternative place-based map (LFshift-PB) aligned the mid-frequency information to cochlear tonotopicity and provided more of the speech spectrum by compressing low-frequency information on the apical electrode contacts (i.e., <1 kHz). Three actual cases of a 12-channel, 24-mm electrode array were simulated by assigning the carrier frequency for an individual channel as the cochlear place frequency of the associated electrode contact. The AID and cochlear place frequency for the most apical electrode contact were 460° and 498 Hz for case 1, 389° and 728 Hz for case 2, and 335° and 987 Hz for case 3, respectively. Results: Generally, better performance was observed with the Strict-PB maps for cases 1 and 2, where mismatches were 2–4 octaves for the most apical channel with the LFshift-PB map. Similar performance was observed between maps for case 3. For the CI simulations, performance with the Strict-PB map declined with decreases in AID, while performance with the LFshift-PB map remained stable across cases. For the EAS simulations, performance with the Strict-PB map remained stable across cases, while performance with the LFshift-PB map improved with decreases in AID. Conclusions: Listeners demonstrated differences with the Strict-PB versus LFshift-PB maps as a function of AID and whether acoustic low-frequency information was available (CI vs. EAS). These data support the use of the Strict-PB mapping procedure for AIDs ≥335°, though further study including time for acclimatization in CI and EAS users is warranted.

An objective comparison of speech intelligibility and quality of hearing aid of different companies

Abstract Purpose: This study aimed to verify the intelligibility and quality of hearing aid-processed speech across companies through objective measures. The output of hearing aids programmed to different prescriptive formulae (Company fit, NAL NL 1, NAL NL 2, and DSLv5) across companies (Company I, Company II, Company III, Company IV, and Company V) were recorded and stored in the sound level meter connected to the manikin. Materials and Methods: Offline analysis of speech quality of the recorded speech was done using the Hearing Aid Speech Quality Index (HASQI), the intelligibilty was measured using the Hearing Aid Speech Perception Index (HASPI) and Speech Intelligibility Index (SII), and Long-Term Average Speech Spectrum (LTASS) was used to measure the spectral composition of speech. Results: When the companies are compared, the hearing aid of Company V showed significantly higher SII values for all prescriptive formulae indicative of better speech intelligibility provided by Company V hearing aids. In terms of speech quality, all the hearing aids showed similar HASQI values. Conclusions: The reasons for each of these findings are highlighted and recommendations for best practices on the use of objective measures for verifying the hearing aid output in clinical setups are suggested.

Improved Speech Spatial Covariance Matrix Estimation for Online Multi-Microphone Speech Enhancement.

Online multi-microphone speech enhancement aims to extract target speech from multiple noisy inputs by exploiting the spatial information as well as the spectro-temporal characteristics with low latency. Acoustic parameters such as the acoustic transfer function and speech and noise spatial covariance matrices (SCMs) should be estimated in a causal manner to enable the online estimation of the clean speech spectra. In this paper, we propose an improved estimator for the speech SCM, which can be parameterized with the speech power spectral density (PSD) and relative transfer function (RTF). Specifically, we adopt the temporal cepstrum smoothing (TCS) scheme to estimate the speech PSD, which is conventionally estimated with temporal smoothing. Furthermore, we propose a novel RTF estimator based on a time difference of arrival (TDoA) estimate obtained by the cross-correlation method. Furthermore, we propose refining the initial estimate of speech SCM by utilizing the estimates for the clean speech spectrum and clean speech power spectrum. The proposed approach showed superior performance in terms of the perceptual evaluation of speech quality (PESQ) scores, extended short-time objective intelligibility (eSTOI), and scale-invariant signal-to-distortion ratio (SISDR) in our experiments on the CHiME-4 database.

Changes in speech and voice characteristics following two hours and eight hours of acute sleep deprivation

Objectives: Sleep loss is known to affect aircrew performance. The study intended to examine the changes in speech and voice characteristics following 2 and 8 h of sleep deprivation for their potential usage as screening methods for detecting sleep deprivation in aircrew. Material and Methods: In a repeated measure design, the study analyzed the effects of 2 h and 8 h of acute sleep deprivation on speech and voice parameters among 30 healthy male volunteers by analyzing standard paragraph reading for 15 s and sustained vowel/a/for 6 s using the commercially available Dr. Speech Voice assessment tool. Subjective appreciation of sleepiness and fatigue was measured by Stanford Sleepiness Scale (SSS) and Chalder’s Fatigue Scale (CFS), respectively, and objective assessment of performance by psychomotor vigilance task (PVT). Results: The subjective markers (SSS and CFS) and objective parameters (PVT reaction time) indicated evidence of fatigue onset even after 2-h of sleep deprivation. In speech parameters, average fundamental frequency (F0), F0 range, and % silence time decreased significantly; whereas, speech time, percent voiceless time, and mean percent voiced time increased significantly following 8-h of sleep deprivation, while percent silence time decreased significantly after 8-h of sleep deprivation . In voice parameters, shimmer % and NNE increased significantly while harmonic-to-noise ratio and signal-to-noise ratio decreased significantly after 8-h of sleep deprivation. Mean shimmer % showed a significant change even after 2-h of sleep deprivation. Conclusion: The findings of the study bring out significant changes in both the speech and voice spectrum following periods of acute sleep deprivation. The changes were more pronounced during sleep deprivation of 8-h as compared to 2-h. These findings have potential application in aviation environment in objective detection of aircrew fatigue following periods of continuous wakefulness and resultant sleep deprivation.

Progressive distribution adapted neural networks for cross-corpus speech emotion recognition.

In this paper, we investigate a challenging but interesting task in the research of speech emotion recognition (SER), i.e., cross-corpus SER. Unlike the conventional SER, the training (source) and testing (target) samples in cross-corpus SER come from different speech corpora, which results in a feature distribution mismatch between them. Hence, the performance of most existing SER methods may sharply decrease. To cope with this problem, we propose a simple yet effective deep transfer learning method called progressive distribution adapted neural networks (PDAN). PDAN employs convolutional neural networks (CNN) as the backbone and the speech spectrum as the inputs to achieve an end-to-end learning framework. More importantly, its basic idea for solving cross-corpus SER is very straightforward, i.e., enhancing the backbone's corpus invariant feature learning ability by incorporating a progressive distribution adapted regularization term into the original loss function to guide the network training. To evaluate the proposed PDAN, extensive cross-corpus SER experiments on speech emotion corpora including EmoDB, eNTERFACE, and CASIA are conducted. Experimental results showed that the proposed PDAN outperforms most well-performing deep and subspace transfer learning methods in dealing with the cross-corpus SER tasks.

과학영재학교 재학생 영어발화 주파수 대역별 음향 에너지 분포의 영어 성취도 예측성 연구*

The average acoustic distribution of American English speakers was statistically compared with the English-speaking patterns of gifted students in a Science Academy in Korea. By analyzing speech recordings, the duration time of which is much longer than in previous studies, this research identified the degree of acoustic proximity between the two parties and the predictability of English academic achievement of gifted high school students. Long-term spectral acoustic power distribution vectors were obtained for 2,048 center frequencies in the range of 20 Hz to 20,000 Hz by applying an long-term average speech spectrum (LTASS) MATLAB code. Three more variables were statistically compared to discover additional indices that can predict future English academic achievement: the receptive vocabulary size test, the cumulative vocabulary scores of English formative assessment, and the English Speaking Proficiency Test scores. Linear regression and correlational analyses between the four variables showed that the receptive vocabulary size test and the low-frequency vocabulary formative assessments which require both lexical and domain-specific science background knowledge are relatively more significant variables than a basic suprasegmental level English fluency in the predictability of gifted students’ academic achievement.

Neural Comb Filtering Using Sliding Window Attention Network for Speech Enhancement

In this paper, we demonstrate the significance of restoring harmonics of the fundamental frequency (pitch) in the deep neural network (DNN)-based speech enhancement. The parameters of the DNN can be estimated by minimizing the mask loss, but it does not restore the pitch harmonics, especially at higher frequencies. In this paper, we propose to restore the pitch harmonics in the spectral domain by minimizing cepstral loss around the pitch peak. Restoring the cepstral pitch peak, in turn, helps in restoring the pitch harmonics in the enhanced spectrum. The proposed cepstral pitch-peak loss acts as an adaptive comb filter on voiced segments and emphasizes the pitch harmonics in the speech spectrum. The network parameters are estimated using a combination of mask loss and cepstral pitch-peak loss. We show that this combination offers the complementary advantages of enhancing both the voiced and unvoiced regions. The DNN-based methods primarily rely on the network architecture, and hence, the prediction accuracy improves with the increasing complexity of the architecture. The lower complex models are essential for real-time processing systems. In this work, we propose a compact model using a sliding-window attention network (SWAN). The SWAN is trained to regress the spectral magnitude mask (SMM) from the noisy speech signal. Our experimental results demonstrate that the proposed approach achieves comparable performance with the state-of-the-art noncausal and causal speech enhancement methods with much lesser computational complexity. Our three-layered noncausal SWAN achieves 2.99 PESQ on the Valentini database with only \(10^9\) floating-point operations (FLOPs).

The Intelligibility of the Reversed-Stethoscope Technique in Age-Related Hearing Loss.

BackgroundThis study evaluated the effectiveness of the reverse stethoscope technique in improving speech intelligibility. In this technique, a clinician places the earpieces of their stethoscope into the ears of a hearing-impaired patient and speaks into the chest piece.MethodsThe International Speech Test Signal was presented to four Littman® stethoscope models and a Pocketalker® personal voice amplifier using an Audioscan® hearing instrument test box. The acoustic outputs of the stethoscopes and voice amplifier were measured across the frequency spectrum of speech. The Speech Intelligibility Index of the resulting speech was calculated for natural speech and for each device in relation to 10 standardized hearing losses representing the population of older adults.ResultsFor each of the 10 hearing losses, the speech signal emitted by the stethoscopes was quieter and yielded lower speech intelligibility scores than regular speech. In contrast, the voice amplifier provided mid- and high-frequency amplification and improved speech intelligibility scores for all but the mildest hearing losses.ConclusionsThe reverse stethoscope technique worsens the clarity of speech and should not be used with older, hearing-impaired patients. Instead, clinicians should use regular speech or, preferably, personal voice amplifiers.

Multi-scale informative perceptual network for monaural speech enhancement

For speech enhancement, contextual information is important for accurate speech spectrum estimation. Conventional convolution layers are generally leveraged to mine implicit correlations from the adjacent area. But non-local information cannot be well captured such as correlations between the pitch and its overtones or full-band noise by fixed convolution. To capture superior dependency along temporal and frequency dimensions, we introduce a multi-scale informative perceptual network (MIPNet) to probe into feature extraction by incorporating localized patterns and global correlations for monaural speech enhancement. MIPNet is based on the encoder-decoder composed of multi-scale perceptual modules (MPMs) to extract preferable local patterns, which have two branches with dilated convolution and stacked fully convolutional layers. MPM is designed with long-term contexts sensitivity to detect the multi-scale adjacent information, thus it helps to rectify informative features and improve the efficiency and accuracy of feature coding. Besides, non-local modules are applied as bottleneck layers to obtain global informative flow. Incorporating MPMs and non-local modules, our proposed network can aggregate multi-scale contextual information, which can model preferable implicit acoustic features and eliminate the noise components. On Voice Bank + DEMAND dataset, MIPNet obtains 14.34% improvement in SSNR for its superiority in noise suppression. Experimental results on WSJ0, TIMIT demonstrate that the proposed model with a few parameters exhibits strong robustness and good performance in terms of objective speech intelligibility and quality under various noise conditions.

Research on Feature Fusion Speech Emotion Recognition Technology for Smart Teaching

At present, there are numerous intelligent teaching applications based on various advanced pattern recognition technologies, such as face sign-in, classroom action recognition, student facial expression recognition, and other systems which have been gradually applied to major schools. Speech emotion recognition can analyze the characteristics of current teaching emotions, so as to discover the rules and details of teachers’ grasp of emotions. In order to realize the task of emotion classification in the process of intelligent teaching through machine learning technology, a teaching speech emotion recognition method with multifeature fusion and deep learning is proposed. The proposed method front-end processes the speech spectrum features in a multifeature fusion manner and combines them with an artificial neural network classifier to form the final speech emotion recognition model. First, after speech preprocessing, three features are selected and fused features using a network structure trained by parallel subnetworks. Then, the classifier used is a hybrid neural network classifier combining a convolutional neural network and a recurrent neural network. Finally, 100 open web courses were used to train and test the model. The test results show that the hybrid neural network sound model using multifeature fusion has good teaching speech emotion recognition capability.

Evaluation of Multimedia Popular Music Teaching Effect Based on Audio Frame Feature Recognition Technology

Music education should pay attention to popular music that exists in students’ real life and deeply affects them. Moreover, it needs to be combined with “popular classical music” to make them happily learn popular music, appreciate popular music artistically, and feel popular music aesthetically. This study combines the audio frame feature recognition technology to evaluate the effect of multimedia popular music teaching and improve the quality of multimedia popular music teaching. Moreover, this study adaptively revises the speech spectrum technology to construct a multimedia pop music system based on audio frame feature recognition technology. Finally, this study verifies the performance of this system through experimental research. According to the results of experimental research, it can be seen that the effect of the system proposed in this study is very good.

The scope and potential of music therapy in stroke rehabilitation

There is a growing interest in the use of music therapy in neurological rehabilitation. Of all the major neurological illnesses, stroke rehabilitation has been observed to have some of the strongest potential for music therapy’s beneficial effect. The current burden of stroke has raised the need to embrace novel, cost-effective, rehabilitation designs that will enhance the existing physical, occupation, and speech therapies. Music therapy addresses a broad spectrum of motor, speech, and cognitive deficits, as well as behavioral and emotional issues. Several music therapy designs have focused on gait, cognitive, and speech rehabilitation, but most of the existing randomized controlled trials based on these interventions have a high risk of bias and are statistically insignificant. More randomized controlled trials with greater number of participants are required to strengthen the current data. Fostering an open and informed dialogue between patients, healthcare providers, and music therapists may help increase quality of life, dispel fallacies, and guide patients to specific musical interventions.

Vector Quantization of Speech Spectrum Based on the VQ-VAE Embedding Space Learning by GAN Technique

Vector Quantization of Speech Spectrum Based on the VQ-VAE Embedding Space Learning by GAN Technique

Speech intelligibility deterioration for normal hearing and hearing impaired patients with different types of tinnitus

ObjectivesMany tinnitus subjects report problems with communication, in particular, difficulties with the intelligibility of speech when it is presented in the background of noise. The type of tinnitus (tone-like, noise, etc.), its location and range in the frequency domain, and the type and degree of accompanying hearing loss can affect speech intelligibility in noise in different ways. The main purpose of this study was to determine the effects of tinnitus and degree of hearing loss on the intelligibility of speech when it is presented in a background noise. MethodsA group of 128 patients participated in the study. There were persons with tinnitus and sensorineural hearing loss as well persons with tinnitus without hearing loss. All participants were patients of the Laryngological Centre of Rehabilitation, Medical University in Poznań. The age of patients ranged from 31 to 84 years. An additional group of 10 subjects (24–50 years) with normal hearing and without tinnitus took part in this study as a control group. An initial experiment was concerned with the determination of the tinnitus type. A band of noise of different widths (Q-factor, goodness), with varying centre frequency (f) was used to match the perceived tinnitus type. The Q factor is the ratio of the centre frequency to the noise bandwidth. In the main experiment, the Speech Reception Threshold (SRT) was measured using The Polish Sentence Test (PST). In this test, short sentences were presented in a background of a so-called babble-noise reflecting the averaged spectrum of Polish speech. Resultsamong the different types of signal used by patients to match their tinnitus, the ones most often used were broadband noise in the medium frequency range (BM) and tone-like high-frequency noise (TH). The average Speech Reception Threshold (SRT) for tinnitus patients with normal hearing was 3 dB higher than that for the control group. The highest deterioration in speech intelligibility was observed for broadband tinnitus, located in the mid-frequency band (1.5–5 kHz). ConclusionsTinnitus patients with normal hearing threshold have significantly higher speech reception thresholds (SRT, on average by 3 dB) than normally hearing subjects (control group, without tinnitus), when the speech is presented in a background of babble-noise. Patients with flat audiograms had SRT values on average about 2.5 dB lower than patients with sloping type audiograms (with greater loss at higher frequencies). On average patients in the age group up to 60 years had 4 dB higher SRT values than the control group, and in the older group of patients (above 61 years) - values were 5 dB higher than in the control group. A significant effect of tinnitus located in the medium and high frequency band on the increase of SRT was noted in normally hearing patients and also in patients with mild hearing loss. For higher degrees of hearing loss, the factor that determines the deterioration of speech intelligibility is the hearing loss (moderate and severe), not tinnitus.

Comparison of discrete transforms for deep‐neural‐networks‐based speech enhancement

In recent studies of speech enhancement, a deep-learning model is trained to predict clean speech spectra from the known noisy spectra of speech. Rather than using the traditional discrete Fourier transform (DFT), this paper considers other well-known transforms to generate the speech spectra for deep-learning-based speech enhancement. In addition to the DFT, seven different transforms were tested: discrete Cosine transform, discrete Sine transform, discrete Haar transform, discrete Hadamard transform, discrete Tchebichef transform, discrete Krawtchouk transform, and discrete Tchebichef-Krawtchouk transform. Two deep-learning architectures were tested: convolutional neural networks (CNN) and fully connected neural networks. Experiments were performed for the NOIZEUS database, and various speech quality and intelligibility measures were adopted for performance evaluation. The quality and intelligibility scores of the enhanced speech demonstrate that discrete Sine transformation is better suited for the front-end processing with a CNN as it outperformed the DFT in this kind of application. The achieved results demonstrate that combining two or more existing transforms could improve the performance in specific conditions. The tested models suggest that we should not assume that the DFT is optimal in front-end processing with deep neural networks (DNNs). On this basis, other discrete transformations should be taken into account when designing robust DNN-based speech processing applications.

Channel and channel subband selection for speaker diarization

Speaker diarization can be considered to be one of the complex problems in speaker recognition. A reliable diarization system should be able to accurately determine the variable length utterances which a speaker contributes to multi-speaker conversations. This is a difficult problem since text-independent speaker identification and verification is yet to be improved for it to be applied reliably. While efficient speaker modelling is important for diarization, the acoustical representation of speech is the basic entity that signifies a speaker. This representation should be outstanding enough to prevent a speaker’s utterances from being lost in the acoustical congestion that is imposed by the rest of the talkers.For this purpose, it is proposed here, for the case of multiple-microphone diarization, multiple speech signals are used in the acoustic feature extraction instead of combining the signals beforehand. The reason is to make an optimal use of those signals in order to enrich the quality of the acoustical representation of the speaker. To this end, and since not all microphone signals (channels) may be desirable, two selection approaches are proposed in this work. These are, a best quality channel selection method and a novel approach for diverse channel selection. Furthermore, a novel method is proposed which retains the speech spectrum from selected least reverberated subbands of the available channels’ spectrums. A new model, referred to here as Averaged Joint Gradient (AJG), is introduced for this purpose. The proposed approach reduces the Diarization Error Rate (DER) in both of the diarization systems used in the evaluations. The first system is based on binary keys and achieves a maximum relative reduction in DER of 14%. The second one is a Gaussian Mixture Model-Bayesian Information Criterion (GMM-BIC) based system which achieves a maximum relative reduction in DER of 20%.

Differentiation of speech-induced artifacts from physiological high gamma activity in intracranial recordings

There is great interest in identifying the neurophysiological underpinnings of speech production. Deep brain stimulation (DBS) surgery is unique in that it allows intracranial recordings from both cortical and subcortical regions in patients who are awake and speaking. The quality of these recordings, however, may be affected to various degrees by mechanical forces resulting from speech itself. Here we describe the presence of speech-induced artifacts in local-field potential (LFP) recordings obtained from mapping electrodes, DBS leads, and cortical electrodes. In addition to expected physiological increases in high gamma (60–200 Hz) activity during speech production, time-frequency analysis in many channels revealed a narrowband gamma component that exhibited a pattern similar to that observed in the speech audio spectrogram. This component was present to different degrees in multiple types of neural recordings. We show that this component tracks the fundamental frequency of the participant's voice, correlates with the power spectrum of speech and has coherence with the produced speech audio. A vibration sensor attached to the stereotactic frame recorded speech-induced vibrations with the same pattern observed in the LFPs. No corresponding component was identified in any neural channel during the listening epoch of a syllable repetition task. These observations demonstrate how speech-induced vibrations can create artifacts in the primary frequency band of interest. Identifying and accounting for these artifacts is crucial for establishing the validity and reproducibility of speech-related data obtained from intracranial recordings during DBS surgery.

New replay attack detection using iterative adaptive inverse filtering and high frequency band

Replay attack is most vulnerable to automatic speaker verification system, where the frauds get the access by replaying pre-recorded voice samples of the genuine speakers. In this proposed work, we mainly concentrated on glottal excitation and high frequency band. First, we demonstrate the importance of the glottal information of the speech signals to detect the replay attack for speaker verification system along with magnitude based discrimination power features set. Iterative Adaptive Inverse Filtering (IA-IF) technique is used to extract the glottal excitation information of the given speech spectrum which shows the difference in the characteristics of genuine and replayed speech sample. Using this technique the glottal information is gained by eliminating the lip radiation and vocal tract effect by applying the integration and inverse filtering respectively. Secondly, we have shown the prominence and discriminative information to detect the replayed attack which is present in high frequency band of the speech spectrums of genuine and replayed samples. Finally, Constant-Q Cepstral Coefficients (CQCC) is used to extract desired features set from the high-frequency bands of glottal excitation spectrum. From the experimental studies done on ASV spoof-2017 version 2.0 database it shows that the proposed method feature set significantly decreases the Equal Error Rate (EER) to 3.68% and 8.32% for development and evaluation set when compare to other state-of-art method feature set.