• 대한전자공학회논문지SD
• /
• 제39권9호
• /
• pp.37-44
• /
• 2002

학습된 신경망(Pre-trained neural network)은 고정된 가중치(weight)를 갖는다. 이 논문에서는 이러한 특성을 이용하여 신경망의 효과적인 디지털 하드웨어의 설계방법을 제안한다. 이를 위해 신경망의 PEs(Processing Elements)연산은 행렬-벡터 곱셈으로 표하고 고정된 가중치와 입력 데이터의 관계를 비트-레벨 어레이(array) 구조로 표현하여, 노드 소거와 가중치 비트 패턴에 따른 공유 노드 설정을 통한 최적화로 연산에 필요한 노드를 최소화한다. FPGA 시뮬레이션 결과, 완전한 정확성에 기반한 하드웨어를 설계하는 경우, 하드웨어 비용을 상당부분 줄였고 동작 주파수가 높다는 것을 확인하였다. 또한, 제안한 설계방법은 한정된 공간 내에서 많은 수의 PEs 구현이 가능함으로, 큰 신경망 모델에 대한 온-칩(on-chip) 구현이 가능하다.

• 한국축산식품학회지
• /
• 제35권5호
• /
• pp.577-584
• /
• 2015

This study was conducted to evaluate the effect of pre-rigor salting level (0-4% NaCl concentration) on physicochemical and textural properties of pre-rigor chicken breast muscles. The pre-rigor chicken breast muscles were de-boned 10 min post-mortem and salted within 25 min post-mortem. An increase in pre-rigor salting level led to the formation of high ultimate pH of chicken breast muscles at post-mortem 24 h. The addition of minimum of 2% NaCl significantly improved water holding capacity, cooking loss, protein solubility, and hardness when compared to the non-salting chicken breast muscle (p<0.05). On the other hand, the increase in pre-rigor salting level caused the inhibition of myofibrillar protein degradation and the acceleration of lipid oxidation. However, the difference in NaCl concentration between 3% and 4% had no great differences in the results of physicochemical and textural properties due to pre-rigor salting effects (p>0.05). Therefore, our study certified the pre-rigor salting effect of chicken breast muscle salted with 2% NaCl when compared to post-rigor muscle salted with equal NaCl concentration, and suggests that the 2% NaCl concentration is minimally required to ensure the definite pre-rigor salting effect on chicken breast muscle.

• 한국염색가공학회지
• /
• 제16권4호
• /
• pp.36-44
• /
• 2004

This research surveys the differences of fabric mechanical properties according to the different looms and the fabric positions with relation to warp and weft weaving tensions on the Picanol and Omega looms respectively. For this purpose, the grey fabrics woven in the previous paper are processed in dyeing and finishing processes, and the processing shrinkages are measured on each processes such as dryer, scouring, pre-set, dyeing and final-set using the fabric density and width. The mechanical properties of the finished fabrics are measured and discussed with relation to the warp and weft weaving tensions of the two looms and the variations of the fabric mechanical properties according to the fabric positions are also discussed. In addition, the fabric thickness according to the fabric positions such as right, left selvedges and center of the fabrics is also measured and discussed with the characteristics of the Picanol and Omega looms.

• International Journal of Computer Science & Network Security
• /
• 제23권1호
• /
• pp.125-139
• /
• 2023

Facial expression recognition, a topical problem in the field of computer vision and pattern recognition, is a direct means of recognizing human emotions and behaviors. This paper first summarizes the datasets commonly used for expression recognition and their associated characteristics and presents traditional machine learning algorithms and their benefits and drawbacks from three key techniques of face expression; image pre-processing, feature extraction, and expression classification. Deep learning-oriented expression recognition methods and various algorithmic framework performances are also analyzed and compared. Finally, the current barriers to facial expression recognition and potential developments are highlighted.

• 한국축산식품학회지
• /
• 제37권1호
• /
• pp.105-113
• /
• 2017

The effect of fermented spinach extracts on color development in cured meats was investigated in this study. The pH values of raw cured meats without addition of fermented spinach extract or nitrite (negative control) were higher (p<0.05) than those added with fermented spinach extract. The pH values of raw and cooked cured meats in treatment groups were decreased with increasing addition levels of fermented spinach extract. The lightness and yellowness values of raw cured meats formulated with fermented spinach extract were higher (p<0.05) than those of the control groups (both positive and negative controls). The redness values of cooked cured meats were increased with increasing fermented spinach extract levels, whereas the yellowness values of cooked cured meats were decreased with increasing levels of fermented spinach extract. The lowest volatile basic nitrogen (VBN) and thiobarbituric acid reactive substances (TBARS) values were observed in the positive control group with addition of nitrite. TBARS values of cured meats added with fermented spinach extract were decreased with increasing levels of fermented spinach extract and VBN values of curing meat with 30% fermented spinach extract was lower than the other treatments. Total viable bacterial counts in cured meats added with fermented spinach extract ranged from 0.34-1.01 Log CFU/g. E. coli and coliform bacteria were not observed in any of the cured meats treated with fermented spinach extracts or nitrite. Residual nitrite contents in treatment groups were increased with increasing levels of fermented spinach extract added. These results demonstrated that fermented spinach could be added to meat products to improve own curing characteristics.

• 말소리와 음성과학
• /
• 제1권2호
• /
• pp.85-94
• /
• 2009

The purpose of this study was to see the effects of pitch and length patterns on spoken word recognition. In Experiment 1, a syllable monitoring task was used to see the effects of pitch and length on the pre-lexical level of spoken word recognition. For both Seoul dialect speakers and Daegu dialect speakers, pitch and length did not affect the syllable detection processes. This result implies that there is little effect of pitch and length in pre-lexical processing. In Experiment 2, a lexical decision task was used to see the effect of pitch and length on the lexical access level of spoken word recognition. In this experiment, word frequency (low and high) as well as pitch and length was manipulated. The results showed that pitch and length information did not play an important role for Seoul dialect speakers, but that it did affect lexical decision processing for Daegu dialect speakers. Pitch and length seem to affect lexical access during the word recognition process of Daegu dialect speakers.

• 한국지반환경공학회 논문집
• /
• 제13권11호
• /
• pp.69-79
• /
• 2012

선지보 네일공법은 터널의 굴착 전에 지반을 보강하므로 기존 지보시스템에 의한 NATM공법에 비해 굴착과정에서 변위를 최소화할 수 있어 안정성과 시공효율을 높일 수 있는 장점이 있다. 본 연구에서는 지하수위가 높고 미고결지반으로 구성된 천층터널에 대해 선지보 네일공법의 적용성을 현장계측관리 사례을 통해 연구하였으며, 지반보강개념에서 선지보 네일공법의 메카니즘과 기존 지보시스템의 메카니즘을 비교 분석하였다. 상하분할 굴착과정에서 굴착단계별 변위증폭이 뚜렷이 관찰되는 기존 지보시스템에 비해 선지보 네일공법은 초기변위 이후 단계굴착별 변위증폭량이 작고 이내 수렴되는 경향을 보여 다단굴착으로 인한 영향이 미미한 것으로 분석되었다. 선지보 네일공법은 종래 NATM공법과 비교할 때 단계굴착 시 응력변화에 대해 민감하고 아칭효과가 발생되지 않는 취약한 지반 즉, 토사터널인 천층터널에서 그 적용성이 뛰어난 것으로 판단된다.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2006년도 Proceedings of ISRS 2006 PORSEC Volume I
• /
• pp.67-70
• /
• 2006

KARI is developing Image Data Acquisition and Control System (IDACS) for pre-processing meteorological and ocean data acquired on geostationary orbit. This paper describes the functions and architecture of IDACS and gives its operation policy including backup operation to overcome limitation of single-configured antenna system. The COMS IDACS provides the capability to receive the raw sensor data and disseminate processed MI data to users via a satellite. From the processed image data, users can produce a set of meteorological and ocean products for a wide range of applications. Most of IDACS subsystems are being developed by Korean technologies and experience acquired from previous projects. In case of COMS geometric correction software module, as it is closely dependent on the characteristics of imagers and spacecraft bus system, it is being co-developed with overseas prime contractor who develops spacecraft bus system.

• 디지털산업정보학회논문지
• /
• 제6권4호
• /
• pp.45-56
• /
• 2010

This paper suggests constraint data modeling based on constraint data presentation techniques to perform complex spatial database operation naturally. We were able to identify the limitation of extendibility of dimension and non-equal framework via relevant research for former schema of spatial database and query processing. Therefore we described generalized tuple of spatial data and the definition of suggested constraint data modeling. Also we selected MLPQ/PReSTO tool among constraint database prototype and compare standard functionality of ARC/VIEW. Then we design scenario for spatial operation using MLPQ/PReSTO and we suggested application effect after query processing. Based on above explanation, we were able to identify that we can process spatial data naturally and effectively using simple constraint routine on same framework via constraint data modeling.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2022년도 학술발표회
• /
• pp.368-368
• /
• 2022

The Soil and Water Assessment Tool (SWAT) has been widely used to simulate the long-term hydrological conditions of a catchment. Two output variables, outflow and sediment yield have been widely investigated in the field of water resources management, especially in determining the conditions of ungauged subbasins. The presence of missing data in weather input data can cause poor representation of the climate conditions in a catchment especially for large or mountainous catchments. Therefore, in this study, a custom module was developed and evaluated to determine the efficiency of utilizing basic spatial interpolation methods in the estimation of weather input data. The module has been written in Python language and can be considered as a pre-processing module prior to using the SWAT model. The results of this study suggests that the utilization of the proposed pre-processing module can improve the simulation results for both outflow and sediment yield in a catchment, even in the presence of missing data.