• The Journal of the Acoustical Society of Korea
• /
• v.23 no.6
• /
• pp.488-496
• /
• 2004

This paper describes a music information retrieval system which uses humming as the key for retrieval Humming is an easy way for the user to input a melody. However, there are several problems with humming that degrade the retrieval of information. One problem is a human factor. Sometimes people do not sing accurately, especially if they are inexperienced or unaccompanied. Another problem arises from signal processing. Therefore, a music information retrieval method should be sufficiently robust to surmount various humming errors and signal processing problems. A retrieval system has to extract pitch from the user's humming. However pitch extraction is not perfect. It often captures half or double pitches. even if the extraction algorithms take the continuity of the pitch into account. Considering these problems. we propose a system that takes multiple pitch candidates into account. In addition to the frequencies of the pitch candidates. the confidence measures obtained from their powers are taken into consideration as well. We also propose the use of an algorithm with three dimensions that is an extension of the conventional DP algorithm, so that multiple pitch candidates can be treated. Moreover in the proposed algorithm. DP paths are changed dynamically to take deltaPitches and IOIratios of input and reference notes into account in order to treat notes being split or unified. We carried out an evaluation experiment to compare the proposed system with a conventional system. From the experiment. the proposed method gave better retrieval performance than the conventional system.

• Proceedings of the IEEK Conference
• /
• 2002.06c
• /
• pp.13-16
• /
• 2002

This paper proposes a method to track hand gesture and to recognize the gesture pattern using Kalman filter and continuous dynamic programming (CDP). The positions of hands are predicted by Kalman filter, and corresponding pixels to the hands are extracted by skin color filter. The center of gravity of the hands is the same as the input pattern vector. The input gesture is then recognized by matching with the reference gesture patterns using CDP. From experimental results to recognize circle shape gesture and intention gestures such as “Come on” and “Bye-bye”, we show the proposed method is feasible to the hand gesture-based human -computer interaction.

• Journal of Korean Society for Atmospheric Environment
• /
• v.11 no.1
• /
• pp.45-56
• /
• 1995

We considered that characteristics of surface ozone continuous and important meteorological parameters from the data measured 7 air quality continuous monitoring stations during 2 years (1990, 1993) in pusan. The diurnal ozone variation showed a primary peak near 1500LST and a secondary peak of the DP(double peaked) pattern. The episode day was defined when an ozone peak higher than 60 ppb was observed at least one station. The frequency of episode day was 100 (298 hours, 69 days). The frequency of the episode day was higher at Meongryundong and Daeyeondong than other sites and highest in August under control of pacific subtropical high. The high temporatant meteorological parameters accompanying the high episode days. The favorable synoptic environment accompanying ozone episode was distributed to 7 different pattern. These pattern can be taken as a nesessary but not an absolute indicator for predicting the occerrence of an episode.

• The Journal of the Acoustical Society of Korea
• /
• v.15 no.4
• /
• pp.43-49
• /
• 1996

This research proposes a system for speaker independent Korean continuous speech recognition with 247 DDD area names using keyword spotting technique. The applied recognition algorithm is the Dynamic Programming Neural Network(DPNN) based on the integration of DP and multi-layer perceptron as model that solves time axis distortion and spectral pattern variation in the speech. To improve performance, we classify word model into keyword model and non-keyword model. We make an experiment on postprocessing procedure for the evaluation of system performance. Experiment results are as follows. The recognition rate of the isolated word is 93.45% in speaker dependent case. The recognition rate of the isolated word is 84.05% in speaker independent case. The recognition rate of simple dialogic sentence in keyword spotting experiment is 77.34% as speaker dependent, and 70.63% as speaker independent.

• KSBB Journal
• /
• v.16 no.4
• /
• pp.398-402
• /
• 2001

Enzymatic hydrolysis of agar was carried out continuously to produce agarooligosaccharides by immobilized agarase in Packed-Bed Reactor. The reactor was constructed using a acryl tube with an internal diameter of 10 mm and a useful height of 140 mm. The Packed-Bed Reactor was 11 mL reactor volume as its length : diameter ratio was 14 : 1. The operation condition of reaction was performed with an 1 g/L agar concentration at 40$^{\circ}C$, 10 mM MOPS buffer(pH 7.0) and with the flow rate 3 mL∼48 mL/h at a dilution rate of 1.09∼5.45 h$\^$-1/. The hydrolysis products was identified DP6, DP4 and DP2 by HPLC. The conversion rate of agar was about 80% and amount of total agarooligosaccharide was 0.88 mg/mL at Packed-Bed Reactor.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.11
• /
• pp.1066-1070
• /
• 2010

In wireless network, energy efficiency is an important design consideration for continuous multimedia streaming service. This paper proposes a new streaming system, called BACASS (Buffer-Aware and Content-Aware Streaming System) that working on the 802.11e U-APSD (Unscheduled Automatic Power Save). The BACASS leads the DP (Doze Period) of U-APSD for improving energy efficiency by utilizing the PSNR based on content-aware and client buffer occupancy that is hinged on a network-aware streaming system using SVC. The simulation results demonstrate the effectiveness of the proposed streaming system.

• Computers and Concrete
• /
• v.7 no.2
• /
• pp.191-202
• /
• 2010

A laboratory investigation was conducted to characterize the constitutive property behavior of Cor-Tuf, an ultra-high-performance composite concrete. Mechanical property tests (hydrostatic compression, unconfined compression (UC), triaxial compression (TXC), unconfined direct pull (DP), uniaxial strain, and uniaxial-strain-load/constant-volumetric-strain tests) were performed on specimens prepared from concrete mixtures with and without steel fibers. From the UC and TXC test results, compression failure surfaces were developed for both sets of specimens. Both failure surfaces exhibited a continuous increase in maximum principal stress difference with increasing confining stress. The DP tests results determined the unconfined tensile strengths of the two mixtures. The tensile strength of each mixture was less than the generally assumed tensile strength for conventional strength concrete, which is 10 percent of the unconfined compressive strength. Both concretes behaved similarly, but Cor-Tuf with steel fibers exhibited slightly greater strength with increased confining pressure, and Cor-Tuf without steel fibers displayed slightly greater compressibility.

• Animal Bioscience
• /
• v.35 no.9
• /
• pp.1340-1350
• /
• 2022

Objective: Luxi Black Head sheep (LBH) is the first crossbreed specialized for meat production and was developed by crossbreeding Black Head Dorper sheep (DP) and Small Tailed Han sheep (STH) in the farming areas of northern China. Research on the genomic variations and selection signatures of LBH caused by continuous artificial selection is of great significance for identifying the genetic mechanisms of important traits of sheep and for the continuous breeding of LBH. Methods: We explored the genetic relationships of LBH, DP, and several Mongolian sheep breeds by constructing phylogenetic tree, principal component analysis and linkage disequilibrium analysis. In addition, we analysed 29 whole genomes of sheep. The genome-wide selection signatures have been scanned with four methods: heterozygosity (H_P), fixation index (F_ST), cross-population extended haplotype homozygosity (XP-EHH) and the nucleotide diversity (𝜃_π) ratio. Results: The genetic relationships analysis showed that LBH appeared to be an independent cluster closer to DP. The candidate signatures of positive selection in sheep genome revealed candidate genes for developmental process (HoxA gene cluster, BCL2L11, TSHR), immunity (CXCL6, CXCL1, SKAP2, PTK6, MST1R), growth (PDGFD, FGF18, SRF, SOCS2), and reproduction (BCAS3, TRIM24, ASTL, FNDC3A). Moreover, two signalling pathways closely related to reproduction, the thyroid hormone signalling pathway and the oxytocin signalling pathway, were detected. Conclusion: The selective sweep analysis of LBH genome revealed candidate genes and signalling pathways associated with developmental process, immunity, growth, and reproduction. Our findings provide a valuable resource for sheep breeding and insight into the mechanisms of artificial selection.

• Corrosion Science and Technology
• /
• v.7 no.1
• /
• pp.1-5
• /
• 2008

The gas atmosphere in continuous annealing and galvanizing lines alters both composition and microstructure of the surface and sub-surface of sheet steel. The formation and morphology of the oxides of alloying elements in High Strength Interstitial Free (HS-IF), Dual Phase (DP) and Transformation-Induced Plasticity (TRIP) steels are strongly influenced by the furnace dew point, and the presence of specific oxide may result in surface defects and bare areas on galvanized sheet products. The present contribution reviews the progress made recently in understanding the selective formation of surface and subsurface oxides during annealing in hot dip galvanizing and conventional continuous annealing lines. It is believed that the surface and sub-surface composition and microstructure have a pronounced influence on galvanized sheet product surface quality. In the present study, it is shown that the understanding of the relevant phenomena requires a combination of precise laboratory-scale simulations of the relevant technological processes and the use of advanced surface analytical tools.

• Proceedings of the IEEK Conference
• /
• 2001.09a
• /
• pp.421-424
• /
• 2001

개별적인 인식기를 하나의 단일 인식 시스템으로 구성하여 음성과 문자를 인식할 수 있는 공용인식시스템의 성능향상을 위해 온라인 필기에서 전역적인 정보를 추출할 수 있는 비트맵 파라미터 추출 방법을 제안하였다. 제안된 방식에서는 고속의 파라미터 추출을 위해 보간법을 이용한 재샘플링 과정 대신에 새로운 시간열을 구성하는 방식을 이용한다. 제안한 비트맵 파라미터를 본 연구실에서 개발한 음성/문자 공용인식 시스템에 적용하기 위하여 67개의 자소를 5상태 10천이 CHMM(Continuous Hidden Markov Model)모델로 구성한 다음 인식알고리즘으로서는 상태단위로 지속 시간 정보를 제어하는 OnePassDP법을 이용하였다. 실험결과, 제안한 방법을 이용한 경우, 자소인식률은 61.3%에서 85.3%로 24%의 인식률 향상을 가져왔으며, 글자인식률은 64.3%에서 82.2%로 17.9%의 인식률 향상을 가져와 제안한 방법의 유효성을 확인할 수 있었다.