• Speech Sciences
• /
• v.11 no.4
• /
• pp.43-52
• /
• 2004

Model parameters in HMM based speech recognition systems are normally estimated using Maximum Likelihood Estimation(MLE). The MLE method is based mainly on the principle of statistical data fitting in terms of increasing the HMM likelihood. The optimality of this training criterion is conditioned on the availability of infinite amount of training data and the correct choice of model. However, in practice, neither of these conditions is satisfied. In this paper, we propose a training algorithm, MCE(Minimum Classification Error), to improve the performance of a speech recognizer detecting mispronunciation of a foreign language. During the conventional MLE(Maximum Likelihood Estimation) training, the model parameters are adjusted to increase the likelihood of the word strings corresponding to the training utterances without taking account of the probability of other possible word strings. In contrast to MLE, the MCE training scheme takes account of possible competing word hypotheses and tries to reduce the probability of incorrect hypotheses. The discriminant training method using MCE shows better recognition results than the MLE method does.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.26 no.5
• /
• pp.74-85
• /
• 2022

The purpose of this paper is to review the applicability of the current standards for the evaluation method of input variables and performance level in seismic performance evaluation by dynamic plastic analysis of the concrete gravity-type dam to which MCE is applied, and to suggest improvements. To this end, a domestic concrete gravity-type dam was selected as a target facility, dynamic plasticity analysis was performed under various conditions, and applicability to input variables such as concrete tensile strength and breaking energy, was reviewed. By analyzing the effect of cracks at the bottom of the gravity dam on the stability of the activity, an improvement plan for the performance level evaluation method required to secure the water storage function was derived. If the proposed improvement plan is applied, it will have the effect of deriving more reasonable evaluation results than the current seismic performance evaluation method to which MCE is applied.

• MALSORI
• /
• no.57
• /
• pp.165-174
• /
• 2006

In this paper we use a training algorithm, MCE (Minimum Classification Error), to improve the performance of a text-independent speaker identification system. The MCE training scheme takes account of possible competing speaker hypotheses and tries to reduce the probability of incorrect hypotheses. Experiments performed on a small set speaker identification task show that the discriminant training method using MCE can reduce identification errors by up to 54% over a baseline system trained using Bayesian adaptation to derive GMM (Gaussian Mixture Models) speaker models from a UBM (Universal Background Model).

• The Journal of the Korea Contents Association
• /
• v.11 no.12
• /
• pp.497-503
• /
• 2011

In this paper, we proposed the MCE as a GMM training method to improve the performance of environmental sounds recognition. We model the environmental sounds data with newly defined misclassification function using the log likelihood of the corresponding class and the log likelihood of the rest classes for discriminative training. The model parameters are estimated with the loss function using GPD(generalized probabilistic descent). For recognition performance comparison, we extracted the 12 degrees features using preprocessing and MFCC(mel-frequency cepstral coefficients) of the 9 kinds of environmental sounds and carry out GMM classification experiments. According to the experimental results, MCE training method showed the best performance by an average of 87.06% with 19 mixtures. This result confirmed us that MCE training method could be effectively used as a GMM training method in environmental sounds recognition.

• The Journal of the Acoustical Society of Korea
• /
• v.16 no.1
• /
• pp.104-109
• /
• 1997

This paper derives a new training alorithm for WHMM using the well-known MCE/GPD method with experimental results on the E-set. The derived algorithm generalizes the conventional adaptive training algorithm for WHMM, which means that HMMs of multiple competing classes can be trained at the same time. The recognition results on the E-set have shown about 15% and 12% improvement for training and test data, respectively.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.44 no.1
• /
• pp.69-80
• /
• 2002

This study aims to develop a potential centrality evaluation model (PCEM) which can evaluate the potential centrality of villages within the catchment boundaries. PCEM is a tool for evaluation of villages\` centralities by the potential centrality index (PCI) developed in this study. For quantification of PCI, total 31 evaluation criteria on the accessibility to other villages and the natural and human environment of the village were introduced. The weighting values of criteria were calculated from the step wise pair-comparision results of 14 specialists in academic fields on rural planning using by AHP (Analytic Hierachy Process) of MCE (multi-criteria evaluation) method. In the results, the weighting values of the spatial accessibility, the natural environments and the human environments were 448, 338 and 214, respectively, among total value being 1,000.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.46 no.4
• /
• pp.76-81
• /
• 2009

We propose an effective emotion recognition algorithm based on the minimum classification error (MCE) incorporating multi-modal system The emotion recognition is performed based on a Gaussian mixture model (GMM) based on MCE method employing on log-likelihood. In particular, the reposed technique is based on the fusion of feature vectors based on voice signal and galvanic skin response (GSR) from the body sensor. The experimental results indicate that performance of the proposal approach based on MCE incorporating the multi-modal system outperforms the conventional approach.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.4C
• /
• pp.207-216
• /
• 2011

Motion compensated extrapolation (MCE) techniques show inferior performance compared to motion compensated interpolation techniques, since only past frames are used in MCE. MCE techniques are used for the reconstruction of corrupted frames, the up-conversion of frame rates and the generation of side information in the distributed video coding system. In this paper, the performance of various MCE techniques are evaluated and an efficient MCE technique using the forward and backward motion estimation is proposed. In the proposed technique, the present frame is extrapolated by averaging two frames which are generated by forward and backward motion estimation respectively. It is shown that the proposed method produces better PSNR results and less blocking phenomena than conventional methods.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.44 no.4
• /
• pp.85-98
• /
• 2002

The purpose of this study is to develop a land use planning model (LUPM) which can be applied to development of rural villages considering their spatial expansion. The LUPM finds out and allocates the new built site required for the improvement of existing villages. in the development of LUPM, CA (cellular automata) and land suitability analysis methods were applied combinedly. The model uses basically numerical data of CIS based on grid data. Agglomerated settlement, as a type of village for simulation, was adopted. Probability of land use change for optimizing development area was calculated by the six criteria: slope. drainage characteristic, direction of slope, as absolute suitability of grid itself, distance from road. distance from stream and distance from the village center, as relative probability by neighborhood cells. Weighting values of these criteria were quantified by AHP (analytic hierarchy process) method, which is one of MCE(multi-criteria evaluation) method. The algorithm of the model was verified by three example villages: an isolation village, a village with horizontal road, and a village with nodal point of cross road

• The Journal of the Acoustical Society of Korea
• /
• v.24 no.1
• /
• pp.58-68
• /
• 2005

In this paper, we propose a scheme to improve discriminative property in the feature compensation method for robust speech recognition under noisy environments. The estimation of noisy speech model used in existing feature compensation methods do not guarantee the computation of posterior probabilities which discriminate reliably among the Gaussian components. Estimation of Posterior probabilities is a crucial step in determining the discriminative factor of the Gaussian models, which in turn determines the intelligibility of the restored speech signals. The proposed scheme employs minimum classification error (MCE) training for estimating the parameters of the noisy speech model. For applying the MCE training, we propose to identify and determine the 'competing components' that are expected to affect the discriminative ability. The proposed method is applied to feature compensation based on parallel combined mixture model (PCMM). The performance is examined over Aurora 2.0 database and over the speech recorded inside a car during real driving conditions. The experimental results show improved recognition performance in both simulated environments and real-life conditions. The result verifies the effectiveness of the proposed scheme for increasing the performance of robust speech recognition systems.