Cepstral Mean Normalization Research Articles

Speech Perception Hash Authentication Algorithm Based on Immittance Spectral Pairs

According to the situation that traditional speech authentication algorithms don’t be appropriated for present speech communication, we proposed a speech authentication algorithm of perceptual hashing based on Immittance Spectral Pairs. It can satisfy the requirement of the efficiency and the robustness for speech authentication. Firstly, the speech signal pre-processing, for framing, adding window, obtained for each speech frame immittance spectral Pairs parameters, constitute an immittance spectral Pairs parameter matrix. Then process cepstral mean and variance normalization for immittance spectral Pairs parameter matrix, cepstral mean and variance normalization can effectively improve the robustness of the Gaussian white noise. And parameter matrix for non-negative matrix factorization. Finally, quantifying the formed weight matrix and getting perceptual hashing sequences.Experiments show that the proposed algorithm has good robustness for content preserving operations, and it doesn’t reduce the efficiency while meeting robustness, it can satisfy the real-time requirement of speech communication.

Environment mismatch compensation using average eigenspace-based methods for robust speech recognition

The performance of speech recognition systems is adversely affected by mismatch between training and test conditions due to environmental factors. In addition to the case of test data from noisy environments, there are scenarios where the training data itself is noisy. In this study, we propose a series of methods for mismatch compensation between training and test environments, based on our "average eigenspace" approach. These methods are also shown to be effective for non-stationary mismatch conditions. An advantage is that there is no need for explicit adaptation data since the method is applied to incoming test data to find the compensatory transform. We evaluate these approaches on two separate corpora which are collected from realistic car environments: CU-Move and UTDrive. Compared with a baseline system incorporating spectral subtraction, highpass filtering and cepstral mean normalization, we obtain a relative word error rate reduction of 17---26 % by applying the proposed techniques. These methods also result in a dimensionality reduction of the feature vectors allowing for a more compact set of acoustic models in the phoneme space, a property important for automatic speech recognition for small footprint size mobile devices such as cell-phone or PDA's which require ASR in diverse environments.

Distant-talking speaker identification by generalized spectral subtraction-based dereverberation and its efficient computation

Abstract Previously, a dereverberation method based on generalized spectral subtraction (GSS) using multi-channel least mean-squares (MCLMS) has been proposed. The results of speech recognition experiments showed that this method achieved a significant improvement over conventional methods. In this paper, we apply this method to distant-talking (far-field) speaker recognition. However, for far-field speech, the GSS-based dereverberation method using clean speech models degrades the speaker recognition performance. This may be because GSS-based dereverberation causes some distortion between clean speech and dereverberant speech. In this paper, we address this problem by training speaker models using dereverberant speech obtained by suppressing reverberation from arbitrary artificial reverberant speech. Furthermore, we propose an efficient computational method for a combination of the likelihood of dereverberant speech using multiple compensation parameter sets. This addresses the problem of determining optimal compensation parameters for GSS. We report the results of a speaker recognition experiment performed on large-scale far-field speech with different reverberant environments to the training environments. The proposed GSS-based dereverberation method achieves a recognition rate of 92.2%, which compares well with conventional cepstral mean normalization with delay-and-sum beamforming using a clean speech model (49.0%) and a reverberant speech model (88.4%). We also compare the proposed method with another dereverberation technique, multi-step linear prediction-based spectral subtraction (MSLP-GSS). The proposed method achieves a better recognition rate than the 90.6% of MSLP-GSS. The use of multiple compensation parameters further improves the speech recognition performance, giving our approach a recognition rate of 93.6%. We implement this method in a real environment using the optimal compensation parameters estimated from an artificial environment. The results show a recognition rate of 87.8% compared with 72.5% for delay-and-sum beamforming using a reverberant speech model.

Single-channel dereverberation by feature mapping using cascade neural networks for robust distant speaker identification and speech recognition

We present a feature enhancement method that uses neural networks (NNs) to map the reverberant feature in a log-melspectral domain to its corresponding anechoic feature. The mapping is done by cascade NNs trained using Cascade2 algorithm with an implementation of segment-based normalization. Experiments using speaker identification (SID) and automatic speech recognition (ASR) systems were conducted to evaluate the method. The experiments of SID system was conducted by using our own simulated and real reverberant datasets, while the CENSREC-4 evaluation framework was used as the evaluation for the ASR system. The proposed method could remarkably improve the performance of both systems by using limited stereo data and low speaker-variant data as the training data. From the evaluation using SID, we reached 26.0% and 34.8% of error rate reduction (ERR) relative to the baseline by using simulated and real data, respectively, by using only one pair of utterances for matched condition cases. Then, by using combined dataset containing 15 pairs of utterances by one speaker from three positions in a room, we could reach 93.7% of average identification rate (three known and two unknown positions), which was 42.2% of ERR relative to the use of cepstral mean normalization (CMN). From the evaluation using ASR, by using 40 pairs of utterances as the NN training data, we could reach 78.4% of ERR relative to the baseline by using simulated utterances by five speakers. Moreover, we could reach 75.4% and 71.6% of ERR relative to the baseline by using real utterances by five speakers and one speaker, respectively.

Robust feature extraction based on an asymmetric level-dependent auditory filterbank and a subband spectrum enhancement technique

In this paper we introduce a robust feature extractor, dubbed as robust compressive gammachirp filterbank cepstral coefficients (RCGCC), based on an asymmetric and level-dependent compressive gammachirp filterbank and a sigmoid shape weighting rule for the enhancement of speech spectra in the auditory domain. The goal of this work is to improve the robustness of speech recognition systems in additive noise and real-time reverberant environments. As a post processing scheme we employ a short-time feature normalization technique called short-time cepstral mean and scale normalization (STCMSN), which, by adjusting the scale and mean of cepstral features, reduces the difference of cepstra between the training and test environments. For performance evaluation, in the context of speech recognition, of the proposed feature extractor we use the standard noisy AURORA-2 connected digit corpus, the meeting recorder digits (MRDs) subset of the AURORA-5 corpus, and the AURORA-4 LVCSR corpus, which represent additive noise, reverberant acoustic conditions and additive noise as well as different microphone channel conditions, respectively. The ETSI advanced front-end (ETSI-AFE), the recently proposed power normalized cepstral coefficients (PNCC), conventional MFCC and PLP features are used for comparison purposes. Experimental speech recognition results demonstrate that the proposed method is robust against both additive and reverberant environments. The proposed method provides comparable results to that of the ETSI-AFE and PNCC on the AURORA-2 as well as AURORA-4 corpora and provides considerable improvements with respect to the other feature extractors on the AURORA-5 corpus.

Histogram Equalization Using Centroids of Fuzzy C-Means of Background Speakers’ Utterances for Majority Voting Based Speaker Identification

이전 연구에서 퍼지 C-Means의 중심 데이터로 이루어진 보조 데이터를 이용한 히스토그램 등화기법을 제안하였다. 보조 데이터를 이용한 히스토그램 등화기법은 사용하는 참조 집합의 크기에 따라 화자 식별 성능에 영향을 받는다. 그러나 인식 시점에서 최적의 파라미터를 찾기는 어렵다. 이 문제를 해결하기 위해 본 논문에서는 화자 식별을 위한 다수 투표 방식에 기반을 둔 보조 데이터를 이용한 히스토그램 등화기법을 제안한다. 다수 투표 기반의 제안한 방법은 여러 종류의 보조 데이터를 이용한 히스토그램 등화기법으로 입력 음성을 분류한다. 본 연구에서 제안한 방법을 CMN(Cepstral Mean Normalization), MVN(Mean and Variance Normalization), HEQ(Histogram Equalization)와 같은 기존의 특징 정규화 방법 및 보조 데이터를 이용한 히스토그램 등화기법과 비교한다. In a previous work, we proposed a novel approach of histogram equalization using a supplement set which is composed of centroids of Fuzzy C-Means of the background utterances. The performance of the proposed method is affected by the size of the supplement set, but it is difficult to find the best size at the point of recognition. In this paper, we propose a histogram equalization using a supplement set for majority voting based speaker identification. The proposed method identifies test utterances using a majority voting on the histogram equalization methods with various sizes of supplement sets. The proposed method is compared with the conventional feature normalization methods such as CMN(Cepstral Mean Normalization), MVN(Mean and Variance Normalization), and HEQ(Histogram Equalization) and the histogram equalization method using a supplement set.

MFCC AND CMN BASED SPEAKER RECOGNITION IN NOISY ENVIRONMENT

The performance of automatic speaker recognition (ASR) system degrades drastically in the presence of noise and other distortions, especially when there is a noise level mismatch between the training and testing environments. This paper explores the problem of speaker recognition in noisy conditions, assuming that speech signals are corrupted by noise. A major problem of most speaker recognition systems is their unsatisfactory performance in noisy environments. In this experimental research, we have studied a combination of Mel Frequency Cepstral Coefficients (MFCC) for feature extraction and Cepstral Mean Normalization (CMN) techniques for speech enhancement. Our system uses a Gaussian Mixture Models (GMM) classifier and is implemented under MATLAB®7 programming environment. The process involves the use of speaker data for both training and testing. The data used for testing is matched up against a speaker model, which is trained with the training data using GMM modeling. Finally, experiments are carried out to test the new model for ASR given limited training data and with differing levels and types of realistic background noise. The results have demonstrated the robustness of the new system.

On Acoustic Emotion Recognition: Compensating for Covariate Shift

Pattern recognition tasks often face the situation that training data are not fully representative of test data. This problem is well-recognized in speech recognition, where methods like cepstral mean normalization (CMN), vocal tract length normalization (VTLN) and maximum likelihood linear regression (MLLR) are used to compensate for channel and speaker differences. Speech emotion recognition (SER) is an important emerging field in human-computer interaction and faces the same data shift problems, a fact which has been generally overlooked in this domain. In this paper, we show that compensating for channel and speaker differences can give significant improvements in SER by modelling these differences as a covariate shift. We employ three algorithms from the domain of transfer learning that apply importance weights (IWs) within a support vector machine classifier to reduce the effects of covariate shift. We test these methods on the FAU Aibo Emotion Corpus, which was used in the Interspeech 2009 Emotion Challenge. It consists of two separate parts recorded independently at different schools; hence the two parts exhibit covariate shift. Results show that the IW methods outperform combined CMN and VTLN and significantly improve on the baseline performance of the Challenge. The best of the three methods also improves significantly on the winning contribution to the Challenge.

An improved model of masking effects for robust speech recognition system

Performance of an automatic speech recognition system drops dramatically in the presence of background noise unlike the human auditory system which is more adept at noisy speech recognition. This paper proposes a novel auditory modeling algorithm which is integrated into the feature extraction front-end for Hidden Markov Model (HMM). The proposed algorithm is named LTFC which simulates properties of the human auditory system and applies it to the speech recognition system to enhance its robustness. It integrates simultaneous masking, temporal masking and cepstral mean and variance normalization into ordinary mel-frequency cepstral coefficients (MFCC) feature extraction algorithm for robust speech recognition. The proposed method sharpens the power spectrum of the signal in both the frequency domain and the time domain. Evaluation tests are carried out on the AURORA2 database. Experimental results show that the word recognition rate using our proposed feature extraction method has been effectively increased.

Performance Evaluation of CMN for Mel-LPC based Speech Recognition in Different Noisy Environments

This study is intended to develop a noise robust distributed speech recognizer for real-world applications by employing Cepstral Mean Normalization (CMN) for robust feature extraction. The main focus of the work is to cope with different noisy environments. To realize this objective, MelLP based speech analysis has been used in speech coding on the linear frequency scale by applying a first-order all-pass filter instead of a unit delay. Mismatch between training and test phases is reduced through robust feature extraction by applying CMN on Mel-LP cepstral coefficients as an effort to reduce additive noise and channel distortion. The performance of the proposed system has been evaluated on test set A of Aurora-2 database which is a subset of TIDigits database contaminated by additive noises and channel effects. The experiment is conducted on four different noisy environments and the baseline performance, that is, for Mel-LPC the average word accuracy has found to be 59.05%. By applying the CMN on Mel-LP cepstral coefficients, the performance has been improved to 68.02%. It is found that CMN performs significantly better for different noisy environments.

Modulation Spectrum Equalization for Improved Robust Speech Recognition

We propose novel approaches for equalizing the modulation spectrum for robust feature extraction in speech recognition. Common to all approaches in that the temporal trajectories of the feature parameters are first transformed into the magnitude modulation spectrum. In spectral histogram equalization (SHE) and two-band spectral histogram equalization (2B-SHE), we equalize the histogram of the modulation spectrum for each utterance to a reference histogram obtained from clean training data, or perform the equalization with two sub-bands on the modulation spectrum. In magnitude ratio equalization (MRE), we define the magnitude ratio of lower to higher modulation frequency components for each utterance, and equalize this to a reference value obtained from clean training data. These approaches can be viewed as temporal filters that are adapted to each testing utterance. Experiments performed on the Aurora 2 and 4 corpora for small and large vocabulary tasks indicate that significant performance improvements are achievable for all noise conditions. We also show that additional improvements can be obtained when these approaches are integrated with cepstral mean and variance normalization (CMVN), histogram equalization (HEQ), higher order cepstral moment normalization (HOCMN), or the advanced front-end (AFE). We analyze and discuss the reasons for these improvements from different viewpoints with different sets of data, including adaptive temporal filtering, noise behavior on the modulation spectrum, phoneme types, and modulation spectrum distance measures.

DWT and LPC based feature extraction methods for isolated word recognition

In this article, new feature extraction methods, which utilize wavelet decomposition and reduced order linear predictive coding (LPC) coefficients, have been proposed for speech recognition. The coefficients have been derived from the speech frames decomposed using discrete wavelet transform. LPC coefficients derived from subband decomposition (abbreviated as WLPC) of speech frame provide better representation than modeling the frame directly. The WLPC coefficients have been further normalized in cepstrum domain to get new set of features denoted as wavelet subband cepstral mean normalized features. The proposed approaches provide effective (better recognition rate), efficient (reduced feature vector dimension), and noise robust features. The performance of these techniques have been evaluated on the TI-46 isolated word database and own created Marathi digits database in a white noise environment using the continuous density hidden Markov model. The experimental results also show the superiority of the proposed techniques over the conventional methods like linear predictive cepstral coefficients, Mel-frequency cepstral coefficients, spectral subtraction, and cepstral mean normalization in presence of additive white Gaussian noise.

Dereverberation and denoising based on generalized spectral subtraction by multi-channel LMS algorithm using a small-scale microphone array

A blind dereverberation method based on power spectral subtraction (SS) using a multi-channel least mean squares algorithm was previously proposed to suppress the reverberant speech without additive noise. The results of isolated word speech recognition experiments showed that this method achieved significant improvements over conventional cepstral mean normalization (CMN) in a reverberant environment. In this paper, we propose a blind dereverberation method based on generalized spectral subtraction (GSS), which has been shown to be effective for noise reduction, instead of power SS. Furthermore, we extend the missing feature theory (MFT), which was initially proposed to enhance the robustness of additive noise, to dereverberation. A one-stage dereverberation and denoising method based on GSS is presented to simultaneously suppress both the additive noise and nonstationary multiplicative noise (reverberation). The proposed dereverberation method based on GSS with MFT is evaluated on a large vocabulary continuous speech recognition task. When the additive noise was absent, the dereverberation method based on GSS with MFT using only 2 microphones achieves a relative word error reduction rate of 11.4 and 32.6% compared to the dereverberation method based on power SS and the conventional CMN, respectively. For the reverberant and noisy speech, the dereverberation and denoising method based on GSS achieves a relative word error reduction rate of 12.8% compared to the conventional CMN with GSS-based additive noise reduction method. We also analyze the effective factors of the compensation parameter estimation for the dereverberation method based on SS, such as the number of channels (the number of microphones), the length of reverberation to be suppressed, and the length of the utterance used for parameter estimation. The experimental results showed that the SS-based method is robust in a variety of reverberant environments for both isolated and continuous speech recognition and under various parameter estimation conditions.

Combination of SPLICE and Feature Normalization for Noise Robust Speech Recognition

It is well-known that the performance of automatic speech recognition (ASR) systems are easily affected by acoustic mismatch between training and testing conditions. This mismatch is often caused by various kinds of environmental noise or distortion. To reduce the effect of mismatch, feature normalization, feature enhancement, model adaptation, etc. have been studied intensively. Cepstral mean normalization (CMN), mean and variance normalization (MVN) and histogram equalization (HEQ) are well-known methods of feature normalization. Stereo-based piecewise linear compensation for environments (SPLICE) is one of the feature enhancement methods. In this paper, we describe how to combine these methods to effectively improve the robustness of ASR systems. In the experiments performed on the Aurora-2 database, a good combination showed a 41% improvement in word error rate over SPLICE only, and a 25% improvement over the conventional combination of SPLICE and CMN.

켑스트럼 정규화와 켑스트럼 거리기반 묵음특징정규화 방법을 이용한 잡음음성 인식

일반적인 음성인식 시스템은 보통 실내 환경에서는 잘 동작하지만 잡음이 존재하는 실제 환경에서는 여러 가지 잡음의 영향으로 그 성능이 급격히 떨어진다. 본 논문에서는 잡음환경에 강인한 음성인식을 위하여 훈련 환경과 실제 환경의 불일치를 줄이기 위한 방법으로 켑스트럼 거리기반 묵음특징 정규화(CSFN: Cepstral distance based SFN) 방법에 켑스트럼 정규화 방법(CMVN:cepstral mean and variance normalization)을 결합한 CSFN-CMVN 방법을 제안하였다. 이 방법은 켑스트럼 특징의 분포 특성의 차이를 나타내는 켑스트럼 유클리디언 거리를 결합하여 음성/묵음 분류에 사용하여 묵음특징을 정규화하는 CSFN 방법에 켑스트럼 정규화 방법을 결합하는 방법이다. Aurora 2.0 DB를 이용한 실험결과, 제안한 CSFN-CMVN은 기존의 대표적인 묵음특징 정규화 방법인 SFN-I 과 비교했을 때 모든 테스트 세트에 대한 평균 단어인식 정확도에서 약 7%의 인식률 향상을 가져옴을 확인하였다. 또한, 기존의 SFN-II, CSFN에 비해서도 약 6%, 5% 향상되었음을 확인 할 수 있어 제안한 방법의 유효성을 확인할 수 있었다. The speech recognition system works well in general indoor environment. However, the recognition performance is dramatically decreased when the system is used in the real environment because of the several noises. In this paper we proposed CSFN-CMVN to improve the recognition performance of the existing CSFN(Cepstral distance based SFN). The CSFN-CMVN method is a combined method of cepstral normalization with CSFN that normalizes silence features using cepstral euclidean distance to classify speech/silence for better performance. From the test results using Aurora 2.0 DB, we could find out that our proposed CSFN-CMVN improves about 7% of more average word accuracy in all the test sets comparing with the typical silence features normalization SFN-I. We can also get improved accuracy of 6% and 5% respectively in compared tests with the conventional SFN-II and CSFN, showing the effectiveness of our proposed method.

Door phone embedded system for voice based user identification and verification platform

This paper describes the architecture of a door phone embedded system with interactive voice response. Because speech technology is not 100% reliable, the emphasis was on parts that have greater impact on overall performance (audio capture, speech recognition and verification, and power consumption). Using an embedded microphone array increases speech recognition effectiveness in very noisy environments. To increase the speech recognition performance, a null grammar with confidence measure support was used. The speaker verification module was also optimized for nosy environments (using the cepstral mean normalization technique and a universal background model).

Combining Speech Enhancement and Cepstral Mean Normalization for LPC Cepstral Coefficients

A mismatch between the training and testing in noisy circumstance often causes a drastic decrease in the performance of speech recognition system. The robust feature coefficients might suppress this sensitivity of mismatch during the recognition stage. In this paper, we investigate the noise robustness of LPC Cepstral Coefficients (LPCC) by using speech enhancement with feature post-processing. At front-end, speech enhancement in the wavelet domain is used to remove noise components from noisy signals. This enhanced processing adopts the combination of discrete wavelet transform (DWT), wavelet packet decomposition (WPD), multi-thresholds processing etc to obtain the estimated speech. The feature post-processing employs cepstral mean normalization (CMN) to compensate the signal distortion and residual noise of enhanced signals in the cepstral domain. The performance of digit speech recognition systems is evaluated under noisy environments based on NOISEX-92 database. The experimental results show that the presented method exhibits performance improvements in the adverse noise environment compared with the previous features.

Distant-Talking Speech Recognition Based on Spectral Subtraction by Multi-Channel LMS Algorithm

SUMMARY We propose a blind dereverberation method based on spectral subtraction using a multi-channel least mean squares (MCLMS) algorithm for distant-talking speech recognition. In a distant-talking environment, the channel impulse response is longer than the short-term spectral analysis window. By treating the late reverberation as additive noise, a noise reduction technique based on spectral subtraction was proposed to estimate the power spectrum of the clean speech using power spectra of the distorted speech and the unknown impulse responses. To estimate the power spectra of the impulse responses, a variable step-size unconstrained MCLMS (VSS-UMCLMS) algorithm for identifying the impulse responses in a time domain is extended to a frequency domain. To reduce the effect of the estimation error of the channel impulse response, we normalize the early reverberation by cepstral mean normalization (CMN) instead of spectral subtraction using the estimated impulse response. Furthermore, our proposed method is combined with conventional delay-andsum beamforming. We conducted recognition experiments on a distorted speech signal simulated by convolving multi-channel impulse responses with clean speech. The proposed method achieved a relative error reduction rate of 22.4% in relation to conventional CMN. By combining the proposed method with beamforming, a relative error reduction rate of 24.5% in relation to the conventional CMN with beamforming was achieved using only an isolated word (with duration of about 0.6s) to estimate the spectrum of the impulse response.

A temporal warped 2D psychoacoustic modeling for robust speech recognition system

Human auditory system performs better than speech recognition system under noisy condition, which leads us to the idea of incorporating the human auditory system into automatic speech recognition engines. In this paper, a hybrid feature extraction method, which utilizes forward masking, backward masking, and lateral inhibition, is incorporated into mel-frequency cepstral coefficients (MFCC). The integration is implemented using a warped 2D psychoacoustic filter. The AURORA2 database is utilized for testing, and the Hidden Markov Model (HMM) is used for recognition. Comparison is made against lateral inhibition (LI), forward masking (FM), cepstral mean and variance normalization (CMVN), the original 2D psychoacoustic filter and the RASTA filter. Experimental results show that the word recognition rate is significantly improved, especially under noisy conditions.

Unsupervised Equalization of Lombard Effect for Speech Recognition in Noisy Adverse Environments

In the presence of environmental noise, speakers tend to adjust their speech production in an effort to preserve intelligible communication. The noise-induced speech adjustments, called Lombard effect (LE), are known to severely impact the accuracy of automatic speech recognition (ASR) systems. The reduced performance results from the mismatch between the ASR acoustic models trained typically on noise-clean neutral (modal) speech and the actual parameters of noisy LE speech. In this study, novel unsupervised frequency domain and cepstral domain equalizations that increase ASR resistance to LE are proposed and incorporated in a recognition scheme employing a codebook of noisy acoustic models. In the frequency domain, short-time speech spectra are transformed towards neutral ASR acoustic models in a maximum-likelihood fashion. Simultaneously, dynamics of cepstral samples are determined from the quantile estimates and normalized to a constant range. A codebook decoding strategy is applied to determine the noisy models best matching the actual mixture of speech and noisy background. The proposed algorithms are evaluated side by side with conventional compensation schemes on connected Czech digits presented in various levels of background car noise. The resulting system provides an absolute word error rate (WER) reduction on 10-dB signal-to-noise ratio data of 8.7% and 37.7% for female neutral and LE speech, respectively, and of 8.7% and 32.8% for male neutral and LE speech, respectively, when compared to the baseline recognizer employing perceptual linear prediction (PLP) coefficients and cepstral mean and variance normalization.