The modified neural net pattern recognition equations were attempted to apply to speech recognition. The proposed method has a dynamic process of self-organization that has been proved to be successful in recognizing a depth perception in stereoscopic vision. This study has shown that the process has also been useful in recognizing human speech. In the processing, input vocal signals are first compared with standard models to measure similarities that are then given to a process of self-organization in neural net equations. The competitive and cooperative processes are conducted among neighboring input similarities, so that only one winner neuron is finally detected. In a comparative study, it showed that the proposed neural networks outperformed the conventional HMM speech recognizer under the same conditions.

Recently a number of social surveys on community response to environmental noises have been conducted to summarize response relationship obtained from different areas. Some problems have been pointed out in comparing the result of surveys using verbal scales with different number of categories. ICBEN (International Commission on Biological Environment of Noise) Team 6 planned a international joint study and constructed comparable standardized noise annoyance scales using the same method. In Korea the survey was conducted in four areas such as Seoul, Taejon, Taegu, Kwangju. About 100 subjects participated in each area approximately. Finally five verbal annoyance were constructed as follows; 1 (Jeonhyu), 2 (Jokm), 3 (Bikyojerk), 4 (Ajoo), 5 (Umcheongnage).

This paper addresses a novel noise-compensation scheme to solve the mismatch problem between training and testing condition for the automatic speech recognition (ASR) system, specifically in car environment. The conventional spectral subtraction schemes rely on the signal-to-noise ratio (SNR) such that attenuation is imposed on that part of the spectrum that appears to have low SNR, and accentuation is made on that part of high SNR. However, these schemes are based on the postulation that the power spectrum of noise is in general at the lower level in magnitude than that of speech. Therefore, while such postulation is adequate for high SNR environment, it is grossly inadequate for low SNR scenarios such as that of car environment. This paper proposes an efficient spectral subtraction scheme focused specifically to low SNR noisy environment by extracting harmonics distinctively in speech spectrum. Representative experiments confirm the superior performance of the proposed method over conventional methods. The experiments are conducted using car noise-corrupted utterances of Aurora2 corpus.

Correlations between acoustic properties and bone density have been investigated in bovine cancellous bone. Speed of sound (SOS), broadband ultrasonic attenuation (BUA), and broadband ultrasonic reflection (BUR) were measured in 10 defatted bovine cancellous bone specimens in vitro. SOS showed a significant correlation with the apparent density of the bone. A comparable correlation was observed between BUA and the apparent density. BUR was rather highly correlated with the apparent density. It was shown that BUR had a weak correlation with BUA and a significant correlation with SOS. This indicates that the parameter BUR can provide important information that may not be contained in BUA and SOS and, therefore, can be useful as an alternative diagnostic parameter of osteoporosis. As expected, a linear combination of all three ultrasonic parameters in a multiple regression model resulted in a significant improvement in predicting the apparent bone density.

We analyzed the measured acoustic field to explore the characteristics of a horizontal coherence in shallow water. Signal spatial coherence data were obtained in the continental shelf off the east coast of Korea using a horizontal line array. The array was deployed on the bottom of 130 m water depth and a sound source was towed at 26 m depth in the source-receiver ranges of 1-13 ㎞. The source transmitted 200 ㎐ pure tone. Topography and temperature profiles along the source track were measured to investigate the relationship between the horizontal coherence and environment variations. The beam bearing disturbance and array signal gain degradation is examined as parameters of horizontal coherence. The results show that the bearing disturbance is about ± 8° and seems to be affected by temporal variations of temperature caused by internal waves. The array signal gains show degradation more than 5㏈ by the temporal and spatial variations of temperature and by the down-sloped topography.

In this paper, a new speech enhancement method using level adapted wavelet packet is presented. First, we propose a level adapted wavelet packet to alleviate a drawback of the conventional node adapted one in noisy environment. Next, we suggest an adaptive noise estimation method at each node on level adapted wavelet packet tree. Then, for more accurate noise component subtraction, we propose a new estimation method of spectral subtraction weight. Finally, we present a modified spectral subtraction method. The proposed method is evaluated on various noise conditions: speech babble noise, F-l6 cockpit noise, factory noise, pink noise, and Volvo car interior noise. For an objective evaluation, the SNR test was performed. Also, spectrogram test and a very simple listening test as a subjective evaluation were performed.