• The Journal of the Acoustical Society of Korea
• /
• v.34 no.4
• /
• pp.316-320
• /
• 2015

In this paper, the pole filtering concept is applied to the Mel-frequency cepstral coefficient (MFCC) feature vectors in the conventional cepstral mean normalization (CMN) and cepstral mean and variance normalization (CMVN) frameworks. Additionally, performance of the cepstral mean and scale normalization (CMSN), which uses scale normalization instead of variance normalization, is evaluated in speech recognition experiments in noisy environments. Because CMN and CMVN are usually performed on a per-utterance basis, in case of short utterance, they have a problem that reliable estimation of the mean and variance is not guaranteed. However, by applying the pole filtering and scale normalization techniques to the feature normalization process, this problem can be relieved. Experimental results using Aurora 2 database (DB) show that feature normalization method combining the pole-filtering and scale normalization yields the best improvements.

• Phonetics and Speech Sciences
• /
• v.12 no.3
• /
• pp.65-71
• /
• 2020

Feature normalization is a method to reduce the effect of environmental mismatch between the training and test conditions through the normalization of statistical characteristics of acoustic feature parameters. It demonstrates excellent performance improvement in the traditional Gaussian mixture model-hidden Markov model (GMM-HMM)-based speech recognition system. However, in a deep neural network (DNN)-based speech recognition system, minimizing the effects of environmental mismatch does not necessarily lead to the best performance improvement. In this paper, we attribute the cause of this phenomenon to information loss due to excessive feature normalization. We investigate whether there is a feature normalization method that maximizes the speech recognition performance by properly reducing the impact of environmental mismatch, while preserving useful information for training acoustic models. To this end, we introduce the mean and exponentiated variance normalization (MEVN), which is a compromise between the mean normalization (MN) and the mean and variance normalization (MVN), and compare the performance of DNN-based speech recognition system in noisy and reverberant environments according to the degree of variance normalization. Experimental results reveal that a slight performance improvement is obtained with the MEVN over the MN and the MVN, depending on the degree of variance normalization.

• Phonetics and Speech Sciences
• /
• v.9 no.2
• /
• pp.103-110
• /
• 2017

The pole filtering concept has been successfully applied to cepstral feature normalization techniques for noise-robust speech recognition. In this paper, it is proposed to apply the pole filtering selectively only to the speech intervals, in order to further improve the recognition performance for short utterances in noisy environments. Experimental results on AURORA 2 task with clean-condition training show that the proposed selectively pole-filtered cepstral mean normalization (SPFCMN) and selectively pole-filtered cepstral mean and variance normalization (SPFCMVN) yield error rate reduction of 38.6% and 45.8%, respectively, compared to the baseline system.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.47 no.6
• /
• pp.15-24
• /
• 2005

The normalization on the characteristics of water level change at the upstream gauging station was attempted according to the operation of drainage sluice gate of the Geumgang estuary dam. The characteristics were normalized by the analysis of water level change and by the linear-regression of the water level data measured at the inner station of Geumgang estuary dam and upstream gauging station. The results of normalization may be referred to the management of Geumgang estuary lake, the operation of pumping and drainage stations in the shore of the lake. The mean response time of water level change on Ibpo, Ganggyeong and Gyuam water level station were 39,81 and 160 minutes, when sluice gate was opened respectively. The mean velocity of surface wave, the mean displacement of water level change, the mean time of water level change and the mean rate of water level change varied largely depending on the location of gauging station and the characteristics of stream section of the water level gauging station.

• The Journal of the Acoustical Society of Korea
• /
• v.24 no.4
• /
• pp.210-216
• /
• 2005

We Propose a new feature normalization scheme based on eigenspace for achieving robust speech recognition. In general, mean and variance normalization (MVN) is Performed in cepstral domain. However, another MVN approach using eigenspace was recently introduced. in that the eigenspace normalization Procedure Performs normalization in a single eigenspace. This Procedure consists of linear PCA matrix feature transformation followed by mean and variance normalization of the transformed cepstral feature. In this method. 39 dimensional feature distribution is represented using only a single eigenspace. However it is observed to be insufficient to represent all data distribution using only a sin91e eigenvector. For more specific representation. we apply unique na independent eigenspaces to cepstra, delta and delta-delta cepstra respectively in this Paper. We also normalize training data in eigenspace and get the model from the normalized training data. Finally. a feature space rotation procedure is introduced to reduce the mismatch of training and test data distribution in noisy condition. As a result, we obtained a substantial recognition improvement over the basic eigenspace normalization.

• Proceedings of the Korean Statistical Society Conference
• /
• 2002.05a
• /
• pp.175-181
• /
• 2002

cDNA microarray experiments permit us to investigate the expression levels of thousands of genes simultaneously and to make it easy to compare gene expression from different populations. However, researchers are asked to be cautious in interpreting the results because of the unexpected sources of variation such as systematic errors from the microarrayer and the difference of cDNA dye intensity. And the scanner itself calculates both of mean and median of the signal and background pixels, so it follows a selection which raw data will be used in analysis. In this paper, we compare the results in each case of using mean and median from the raw data and normalization methods in reducing the systematic errors with arm's skin cells of old and young males. Using median is preferable to mean because the distribution of the test statistic (t-statistic) from the median is more close to normal distribution than that from mean. Scaled print tip normalization is better than global or lowess normalization due to the distribution of the test-statistic.

• The Journal of the Acoustical Society of Korea
• /
• v.33 no.1
• /
• pp.68-74
• /
• 2014

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.

• The Journal of the Acoustical Society of Korea
• /
• v.35 no.4
• /
• pp.280-287
• /
• 2016

In this paper, we propose a normalization algorithm that can be applied to adaptive filters for multi-channel active noise control. The FxLMS (Filtered-x Least Mean Square) algorithm for the single-channel active noise control can be normalized in the same way as the NLMS (Normalized Least Mean Square) algorithm, whereas in case of the multi-channel active noise control, the single-channel normalization for the FxLMS algorithm cannot be extended to the normalization for the multi-channel FxLMS algorithm straightforwardly. First, we adopt a generalized normalization algorithm for the multi-channel FxLMS algorithm based on the principle of minimal disturbance and then, proposed a normalized algorithm considering only diagonal elements to avoid computation for matrix inversion. We carried out performance comparisons of the proposed algorithm with other algorithms without normalization. It is shown that the proposed algorithm presents better convergence characteristics under non-stationary environments.

• Proceedings of the IEEK Conference
• /
• 2009.05a
• /
• pp.193-195
• /
• 2009

In this paper, we propose an effective compression method for electrocardiogram(ECG) signals. 1-D ECG signals are reconstructed to 2-D ECG data by period and complexity sorting schemes with image compression techniques to Increase inter and intra-beat correlation. The proposed method added block division and mean-period normalization techniques on top of conventional 2-D data ECG compression methods. JPEG 2000 is chosen for compression of 2-D ECG data. Standard MIT-BIH arrhythmia database is used for evaluation and experiment. The results show that the proposed method outperforms compared to the most recent literature especially in case of high compression rate.

• Genomics & Informatics
• /
• v.2 no.2
• /
• pp.92-98
• /
• 2004

A cDNA microarray experiment is one of the most useful high-throughput experiments in medical informatics for monitoring gene expression levels. Statistical analysis with a cDNA microarray medical data requires a normalization procedure to reduce the systematic errors that are impossible to control by the experimental conditions. Despite the variety of normalization methods, this. paper suggests a more general and synthetic normalization algorithm with a control gene set based on previous studies of normalization. Iterative normalization method was used to select and include a new control gene set among the whole genes iteratively at every step of the normalization calculation initiated with the housekeeping genes. The objective of this iterative normalization was to maintain the pattern of the original data and to keep the gene expression levels stable. Spatial plots, M&A (ratio and average values of the intensity) plots and box plots showed a convergence to zero of the mean across all genes graphically after applying our iterative normalization. The practicability of the algorithm was demonstrated by applying our method to the data for the human photo aging study.