• Journal of the Korea Institute of Military Science and Technology
• /
• v.15 no.5
• /
• pp.601-614
• /
• 2012

Since the optimized use of sonar systems available for detection is a very practical problem for a given ocean environment, the measure of mission achievability is needed for operating the sonar system efficiently. In this paper, a theory on Measure Of Effectiveness(MOE) for specific mission such as detection is described as the measure of mission achievability, and a recursive Cumulative Detection Probability(CDP) algorithm is found to be most efficient from comparing three CDP algorithms for discrete glimpses search to reduce computation time and memory for complicated scenarios. The three CDPs which are MOE for sonar-maneuver pattern are calculated as time evolves for comparison, based on three different formula depending on the assumptions as follows; dependent or independent glimpses, unimodal or non-unimodal distribution of Probability of Detection(PD) as a function of observation time interval for detection. The proposed CDP algorithm which is made from unimodal formula is verified and applied to OASPP(Optimal Acoustic Search Path Planning) with complicated scenarios.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.1
• /
• pp.53-60
• /
• 1999

This paper is a study on the korean broadcasting speech recognition. Here we present the methods for the large vocabuary continuous speech recognition. Our main concerns are the language modeling and the search algorithm. The used acoustic model is the uni-phone semi-continuous hidden markov model and the used linguistic model is the N-gram model. The search algorithm consist of three phases in order to utilize all available acoustic and linguistic information. First, we use the forward Viterbi beam search to find word end frames and to estimate related scores. Second, we use the backword Viterbi beam search to find word begin frames and to estimate related scores. Finally, we use A/sup */ search to combine the above two results with the N-grams language model and to get recognition results. Using these methods maximum 96.0% word recognition rate and 99.2% syllable recognition rate are achieved for the speaker-independent continuous speech recognition problem with about 12,000 vocabulary size.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.2E
• /
• pp.10-17
• /
• 1999

The aim of the work described in this paper is to develop a complex underground acoustic system which detects and localizes the origin of an underground hammering sound using an array of hydrophones located about 100m underground. Three different methods for the sound localization will be presented, a time-delay method, a power-attenuation method and a hybrid method. In the time-delay method, the cross correlation of the signals received from the array of sensors is used to calculate the time delays between those signals. In the power-attenuation method, the powers of the received signals provide a measure of the distances of the source from the sensors. In the hybrid method, both informations of time-delays and power-ratios are coupled together to produce better performance of position estimation. A new acoustic imaging technique has been developed for improving the hybrid method. This new acoustic imaging method shows the multi-dimensional distribution of the normalized cost function, so as to indicate the trend of the minimizing direction toward the source location. For each method the sound localization is carried out in three dimensions underground. The distance between the true and estimated origins of the source is 28m for a search area of radius 250m.

• MALSORI
• /
• no.51
• /
• pp.151-165
• /
• 2004

For educational and research purposes, a Korean speech recognition platform is designed. It is based on an object-oriented architecture and can be easily modified so that researchers can readily evaluate the performance of a recognition algorithm of interest. This platform will save development time for many who are interested in speech recognition. The platform includes the following modules: Noise reduction, end-point detection, met-frequency cepstral coefficient (MFCC) and perceptually linear prediction (PLP)-based feature extraction, hidden Markov model (HMM)-based acoustic modeling, n-gram language modeling, n-best search, and Korean language processing. The decoder of the platform can handle both lexical search trees for large vocabulary speech recognition and finite-state networks for small-to-medium vocabulary speech recognition. It performs word-dependent n-best search algorithm with a bigram language model in the first forward search stage and then extracts a word lattice and restores each lattice path with a trigram language model in the second stage.

• Speech Sciences
• /
• v.10 no.1
• /
• pp.71-84
• /
• 2003

In the acoustic modeling for large vocabulary speech recognition, a sparse data problem caused by a huge number of context-dependent (CD) models usually leads the estimated models to being unreliable. In this paper, we develop a new clustering method based on the C45 decision-tree learning algorithm that effectively encapsulates the CD modeling. The proposed scheme essentially constructs a supervised decision rule and applies over the pre-clustered triphones using the C45 algorithm, which is known to effectively search through the attributes of the training instances and extract the attribute that best separates the given examples. In particular, the data driven method is used as a clustering algorithm while its result is used as the learning target of the C45 algorithm. This scheme has been shown to be effective particularly over the database of low unknown-context ratio in terms of recognition performance. For speaker-independent, task-independent continuous speech recognition task, the proposed method reduced the percent accuracy WER by 3.93% compared to the existing rule-based methods.

• Transactions on Control, Automation and Systems Engineering
• /
• v.2 no.1
• /
• pp.69-75
• /
• 2000

The aim of the work described in this paper is to develop a complex underground acoustic system which detects and localizes the origin of an underground hammering sound using an array of hydrophones located about 100m underground. Three different methods for the sound localization will be presented, a time-delay method, a power-attenuation method and a hybrid method. In the time-delay method, the cross correlation of the signals received from the array of sensors is used to calculate the time delays between those signals. In the power-attenuation method, the powers of the received signals provide a measure of the distances of the source from the sensors. In the hybrid method, both informations of time-delays and power-ratios are coupled together to produce better performance of position estimation. A new acoustic imaging technique has been developed for improving the hybrid method. This new acoustic imaging method shows the multi-dimensional distribution of the normalized cost function, so as to indicate the trend of the minimizing direction toward the source location. For each method the sound localization is carried out in three dimensions underground. The distance between the true and estimated origins of the source is 28m for a search area of radius 250m.

• Proceedings of the Acoustical Society of Korea Conference
• /
• spring
• /
• pp.103-108
• /
• 1999

The aim of the work described in this paper is to develop a complex underground acoustic system which detects and localizes the origin of an underground hammering sound using an array of hydrophones located about loom underground. Three different methods for the sound localization will be presented, a time-delay method, a power-attenuation method and a hybrid method. In the time-delay method, the cross correlation of the signals received from the array of sensors is used to calculate the time delays between those signals. In the power-attenuation method, the powers of the received signals provide a measure of the distances of the source from the sensors. In the hybrid method, both informations of time-delays and power-ratios are coupled together to produce better performance of position estimation. A new acoustic imaging technique has been developed for improving the hybrid method. This new acoustic imaging method shows the multi-dimensional distribution of the normalized cost function, so as to indicate the trend of the minimizing direction toward the source location. For each method the sound localization is carried out in three dimensions underground. The distance between the true and estimated origins of the source is 28m for a search area of radius 250m.

• Proceedings of the Korea Committee for Ocean Resources and Engineering Conference
• /
• 2003.05a
• /
• pp.109-114
• /
• 2003

Wireless communication systems are inevitable for efficient underwater activities. Because of the poor propagation characteristics of light and electromagnetic waves, acoustic waves are generally used for the underwater wireless communication. Although there are many kinds of information type, visual images take an essential role especially for search and identification activities. For this reason, we developed an acoustic-based underwater image transmission system under a dual use technology project supported by MOCIE (Ministry of Commerce, Industry and Energy). For the application to complicated and time-varying underwater environments all-digital transmitter and receiver systems are investigated. Array acoustic transducers are used at the receiver, which have the center frequency of 32kHz and the bandwidth of 4kHz. To improve transmission speed and quality, various algorithms and systems are used. The system design techniques will be discussed in detail including image compression/ decompression system, adaptive beam- forming, fast RLS adaptive equalizer, ${\partial}/4$ QPSK (Quadrilateral Phase Shift Keying) modulator/demodulator, and convolution coding/ Viterbi. Decoding.

• The Journal of the Acoustical Society of Korea
• /
• v.26 no.3E
• /
• pp.69-73
• /
• 2007

This paper presents an efficient algorithm to retrieve similar music files from a large archive of digital music database. Users are able to navigate and discover new music files which sound similar to a given query music file by searching for the archive. Since most of the methods for finding similar music files from a large database requires on computing the distance between a given query music file and every music file in the database, they are very time-consuming procedures. By measuring the acoustic distance between the pre-classified music files with the same type of emotion, the proposed method significantly speeds up the search process and increases the precision in comparison with the brute-force method.

• 한국HCI학회:학술대회논문집
• /
• 2008.02a
• /
• pp.360-365
• /
• 2008

We propose an efficient robot name registration and recognition method in order to enable a Call-and-Come service for home robots. In the proposed method for the name registration, the search space is first restricted by using monophone-based acoustic models. Second, the registration of robot names is completed by using triphone-based acoustic models in the restricted search space. Next, the parameter for the utterance verification is calculated to reduce the acceptance rate of false calls. In addition, acoustic models are adapted by using a distance speech database to improve the performance of distance speech recognition, Moreover, the location of a user is estimated by using a microphone array. The experimental result on the registration and recognition of robot names shows that the word accuracy of speech recognition is 98.3%.