• Journal of Korean Institute of Industrial Engineers
• /
• 제40권1호
• /
• pp.140-150
• /
• 2014

University ranking models, though they consider multiple indicators to evaluate universities, determine the overall score of each university based on their own normalization and aggregation methods. Thus, the rankings provided by such models primarily depend on actual scores of evaluation indicators, but they are also significantly affected by the normalization and aggregation methods. We examine the normalization methods of four university ranking models used in South Korea, China, and United Kingdom. We discuss a possible unintended consequence of these methods, i.e., some universities who want to improve their rankings may focus on unnecessary indicators, contrary to the evaluator's intension, due to the normalization methods. We suggest a new normalization method based on the statistical characteristics of the distribution of each evaluation indicator so that the new method can motivate the universities to move into the desirable directions intended by the evaluator.

• Journal of Internet Computing and Services
• /
• 제4권3호
• /
• pp.9-14
• /
• 2003

The method of vocal tract normalization is a successful method for improving the accuracy of inter-speaker normalization. In this paper, we present an intra-speaker warping factor estimation based on pitch alteration utterance. The feature space distributions of untransformed speech from the pitch alteration utterance of intra-speaker would vary due to the acoustic differences of speech produced by glottis and vocal tract. The variation of utterance is two types: frequency and amplitude variation. The vocal tract normalization is frequency normalization among inter-speaker normalization methods. Therefore, we have to consider amplitude variation, and it may be possible to determine the amplitude warping factor by calculating the inverse ratio of input to reference pitch. k, the recognition results, the error rate is reduced from 0.4% to 2.3% for digit and word decoding.

• Phonetics and Speech Sciences
• /
• 제12권3호
• /
• 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.

• Speech Sciences
• /
• 제13권1호
• /
• pp.129-139
• /
• 2006

In this paper, we propose two effective energy feature normalization methods for robust speech recognition in noisy environments. In the first method, we estimate the noise energy and remove it from the noisy speech energy. In the second method, we propose a modified algorithm for the Log-energy Dynamic Range Normalization (ERN) method. In the ERN method, the log energy of the training data in a clean environment is transformed into the log energy in noisy environments. If the minimum log energy of the test data is outside of a pre-defined range, the log energy of the test data is also transformed. Since the ERN method has several weaknesses, we propose a modified transform scheme designed to reduce the residual mismatch that it produces. In the evaluation conducted on the Aurora2.0 database, we obtained a significant performance improvement.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• 제36권8C호
• /
• pp.505-515
• /
• 2011

This paper describes a normalization method of Haar-like features used for object detection. Previous method which performs variance normalization on Haar-like features requires a lot of calculations, since it uses an additional integral image for calculating the standard deviation of intensities of pixels in a candidate window and increases possibility of false detection in the area where variance of brightness is small. The proposed normalization method can be performed much faster than the previous method by not using additional integral image and classifiers which are trained with the proposed normalization method show robust performance in various lighting conditions. Experimental result shows that the object detector which uses the proposed method is 26% faster than the one which uses the previous method. Detection rate is also improved by 5% without increasing false alarm rate and 45% for the samples whose brightness varies significantly.

• The Journal of the Acoustical Society of Korea
• /
• 제34권4호
• /
• 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.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• 제31B권5호
• /
• pp.162-167
• /
• 1994

The binary image normalization method in image processing can be used in several fields, Especially, its high speed processing method and its hardware implmentation is more useful, A normalization process of each character in character recognition requires a lot of processing time. Therefore, the research was done as a part of high speed process of OCR (optical character reader) implementation as a pipeline structure with host computer in hardware to give temporal parallism. For normalization process, general purpose CPU,MC68000, was used to implement it. As a result of experiment, the normalization speed of the hardware is sufficient to implement high speed OCR which the recognition speed is over 140 characters per second.

• Journal of Information Processing Systems
• /
• 제11권4호
• /
• pp.573-582
• /
• 2015

Edit distance metrics are widely used for many applications such as string comparison and spelling error corrections. Hamming distance is a metric for two equal length strings and Damerau-Levenshtein distance is a well-known metrics for making spelling corrections through string-to-string comparison. Previous distance metrics seems to be appropriate for alphabetic languages like English and European languages. However, the conventional edit distance criterion is not the best method for agglutinative languages like Korean. The reason is that two or more letter units make a Korean character, which is called as a syllable. This mechanism of syllable-based word construction in the Korean language causes an edit distance calculation to be inefficient. As such, we have explored a new edit distance method by using consonant normalization and the normalization factor.

• Journal of Internet Computing and Services
• /
• 제19권3호
• /
• pp.35-42
• /
• 2018

Deep learning is known as a method of high accuracy among several methods for image classification. In this paper, we propose a method of enhancing the accuracy of image classification using CNN with a batch normalization method for classification of images using deep CNN (Convolutional Neural Network). In this paper, we propose a method to add a batch normalization layer to existing neural networks to enhance the accuracy of image classification. Batch normalization is a method to calculate and move the average and variance of each batch for reducing the deflection in each layer. In order to prove the superiority of the proposed method, Accuracy and mAP are measured by image classification experiments using five image data sets SHREC13, MNIST, SVHN, CIFAR-10, and CIFAR-100. Experimental results showed that the CNN with batch normalization is better classification accuracy and mAP rather than using the conventional CNN.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• 제39A권3호
• /
• pp.155-157
• /
• 2014

In this paper, a novel codebook generation strategy is proposed. With the given codebooks, two codeword selection procedures are proposed and analyzed for generating the quantized multiple-input multiple-output (MIMO) channel state information (CSI). Furthermore, three different quantization and normalization strategies are analyzed. The simulation results suggest that the proposed 'quantized channel generation method 2' is the best strategy to reduce the quantization and normalization errors to generate the final quantized MIMO CSI.