• Proceedings of the Korea Information Processing Society Conference
• /
• 2007.05a
• /
• pp.397-398
• /
• 2007

본 논문에서는 환경변화에 대해 강인하게 동작하는 음성인식 시스템을 위해 잡음적응 훈련과 변별학습 방식을 결합한 형태의 환경적응 방식을 제안한다. 다중환경 훈련과 잡음제거방식을 결합한 형태인 잡음적응 훈련 방식은 음성인식을 위한 MCE (Minimum Classification Error)의 목적과는 거리가 있고, 음성인식 시스템이 사용되는 모든 환경을 반영하는 것은 현실적으로 어렵다는 점에서 한계가 있다. 이에 잡음적응 훈련방식으로 훈련된 기본 음향모델을 목적환경에서 수집한 소량의 데이터를 이용한 변별학습을 통해 환경적응 모델로 변환함으로써 이러한 단점을 보완할 수 있는 잡음 적응 변별학습을 이용한 훈련방식을 제안한다.

• The Journal of the Acoustical Society of Korea
• /
• v.20 no.7
• /
• pp.47-54
• /
• 2001

In this paper, we present a variable dimension vector quantization for spectral magnitudes. Espectially, spectral magnitudes of the Harmonic coder, need variable dimension quantizer because those are not fixed dimension. So, this paper present efficient quantization methods. These methods use variable Discrete Cosine Transform(DCT) for spectral magnitude parameters and NSTVQ which is combined odd/even, split and multi-stage structure, proposed quantization methods use Spectral Distortion(SD) for performance measure. Consequently, Multi-Stage Nonsquare Transform Vector Quantization(MSNSTVQ) is the best in performance measure.

• The Journal of the Acoustical Society of Korea
• /
• v.30 no.3
• /
• pp.149-157
• /
• 2011

In this paper, we propose a new cardiac disorder classification method using an support vector machine (SVM) to combine hidden Markov model (HMM) and murmur existence information. Using cepstral features and the HMM Viterbi algorithm, we segment input heart sound signals into HMM states for each cardiac disorder model and compute log-likelihood (score) for every state in the model. To exploit the temporal position characteristics of murmur signals, we divide the input signals into two subbands and compute murmur probability of every subband of each frame, and obtain the murmur score for each state by using the state segmentation information obtained from the Viterbi algorithm. With an input vector containing the HMM state scores and the murmur scores for all cardiac disorder models, SVM finally decides the cardiac disorder category. In cardiac disorder classification experimental results, the proposed method shows the relatively improvement rate of 20.4 % compared to the HMM-based classifier with the conventional cepstral features.

• The Journal of the Acoustical Society of Korea
• /
• v.42 no.3
• /
• pp.214-220
• /
• 2023

In this paper, we propose a method to improve texture recognition performance from haptic acceleration signals representing the texture characteristics of object surface materials by using a Conformer model that combines the advantages of a convolutional neural network and a transformer. In the proposed method, three-axis acceleration signals generated by impact sound and vibration are combined into one-dimensional acceleration data while a person contacts the surface of the object materials using a tool such as a stylus , and the logarithmic Mel-spectrogram is extracted from the haptic acceleration signal similar to the audio signal. Then, Conformer is applied to the extracted the logarithmic Mel-spectrogram to learn main local and global frequency features in recognizing the texture of various object materials. Experiments on the Lehrstuhl für Medientechnik (LMT) haptic texture dataset consisting of 60 materials to evaluate the performance of the proposed model showed that the proposed method can effectively recognize the texture of the object surface material better than the existing methods.

• The Journal of the Acoustical Society of Korea
• /
• v.27 no.8
• /
• pp.435-442
• /
• 2008

The n-gram model is appropriate for languages, such as English, in which the word-order is grammatically rigid. However, it is not suitable for Korean in which the word-order is relatively free. Previous work proposed a twoply HMM that reflected the characteristics of Korean but failed to reflect word-order structures among words. In this paper, we define a new segment unit which combines two words in order to reflect the characteristic of word-order among adjacent words that appear in verbal morphemes. Moreover, we propose a two-path language model that estimates probabilities depending on the context based on the proposed segment unit. Experimental results show that the proposed two-path language model yields 25.68% perplexity improvement compared to the previous Korean language models and reduces 94.03% perplexity for the prediction of verbal morphemes where words are combined.

• The Journal of the Acoustical Society of Korea
• /
• v.28 no.4
• /
• pp.308-317
• /
• 2009

In the present work, a sensor system composed of an oscillator sensor and a Lamb wave sensor is proposed for the purpose of structural health monitoring. The oscillator sensing system detecting the shift of a structural resonant frequency in proportion to the amount of defects in the structure is a pretty sensitive and simple device, but its detectable range is limited to its local zone. The Lamb wave sensor system, however, is applicable to global detection of the defects. This study is aimed at investigating the feasible combination of the two systems to exploits their merits simultaneously. The scheme to use PZT patches as the oscillator sensor as well as the Lamb wave sensor was proposed to identify the position, length and number of cracks by means of TOF and amplitude of signals, and its validity was confirmed through experiments.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.4
• /
• pp.60-64
• /
• 1999

In this paper, we apply CART(Classification And Regression tree) and OC1(Oblique Classifier1) which methods are widely used for continuous speech recognition and synthesis. We prediet prosodic boundary index by applying CART and OC1, which combine right depth of tree-structured method and To_Right of link grammar method with tri_gram model. We assigned four prosodic boundary index level from 0 to 3. Experimental results show that OC1 method is superior to CART method. In other words, in spite of OC1's having fewer nodes than CART, it can make more improved prediction than CART.

• The Journal of the Acoustical Society of Korea
• /
• v.11 no.3
• /
• pp.25-39
• /
• 1992

Undesired vibratory motion of a simply supported plate is controlled with piezoelectric sensors and actuators. Appropriate dynamic equations of the sensor and actuator are derived and coupled with the dynamic equation of the plate for the construction of an active feedback vibration control system. Analytic solutions are obtained for amplitude response of the plate, reflecting the combined effect of external driving forces and piezoelectric control moments. Numerical examples are presented to illustrate the effectiveness of this approach for two types of external forces, i.e. a concentrated point load and a piezoelectric plate driver. Calculation results show that the sensors and actuators can be efficient tools to mitigate the sensitivity of the structure to external sources of vibration. The method investigated in this work is applicable to arbitrary external loading conditions and control algorithms.

• The Journal of the Acoustical Society of Korea
• /
• v.32 no.6
• /
• pp.541-547
• /
• 2013

The speech quality of multi-party audio telephony between two, three or more participants is decreased by audio volume imbalance, audio volume saturation and noise level increase. To solve this issue, this paper proposes an advanced audio mixing algorithm for software-based multi-point control unit. Our approach is based on the combined voice activity detection and gain control technique that consists of a set of algorithms that classify audio signals, estimate audio volumes, adjust gain factors and mix audio signals of all channels. The proposed audio mixing algorithm is computationally efficient, delivers high-quality speech, and is suitable for use in any practical multi-party audio telephony.

• The Journal of the Acoustical Society of Korea
• /
• v.38 no.5
• /
• pp.574-579
• /
• 2019

Snoring is a typical symptom of sleep disorder and it is important to identify the occurrence of snoring because it causes sleep apnea. In this paper, we proposes a residual convolutional neural network as an efficient snoring identification algorithm. Residual convolutional neural network, which is a structure combining residual learning and convolutional neural network, effectively extracts features existing in data more than conventional neural network and improves the accuracy of snoring identification. Experimental results show that the performance of the proposed snoring algorithm is superior to that of the conventional methods.