단어, 그림 및 추상패턴의 일화적 부호화 기전: event-related fMRI 연구

박태진,박선희,김정희 한국인지및생물심리학회 2008 한국심리학회지 인지 및 생물 Vol.20 No.3

To investigate how the mechanisms of memory encoding were influenced by the verbalizability of nonverbal material and the imageability of verbal material, an event-related fMRI study was performed. Subjects were asked to intentionally memorize the mixtures of concrete words, object pictures, and abstract patterns and then perform recognition judgement test. Recognition scores revealed the superior accuracy for pictures and the worst accuracy for abstract patterns. Analyses of signal change ratio at the superior PFC(BA 6), superior parietal lobe(BA 7), and medial temporal lobe revealed right-lateralized frontal activation during abstract pattern encoding, no lateralized frontal activation during word and picture encoding, and higher right-frontal activation during abstract pattern encoding than word and picture encoding. No lateralized parietal activation during encoding was observed on all types of learning materials, and higher bilateral-parietal activation during abstract pattern encoding than during word and picture encoding was observed. No lateralized, nor differential MTL activation during encoding was observed on all types of learning materials. Analyses of contrast among learning materials revealed similar activation patterns at superior PFC and superior parietal lobe but different activation patterns at MTL, showing higher bilateral-MTL activation during word and picture encoding than during abstract pattern encoding as well as higher right-MTL activation during picture encoding than during word encoding. The overall bilateral activation patterns of words were similar to those of pictures, but activation patterns of words and pictures were different from those of abstract patterns. The data support the code-specificity rather than material-specificity hypothesis and indicate that the neural mechanisms of intentional memory encoding were dependent on the types of learning materials. 언어적 자료의 심상화 특성과 비언어적 자료의 언어화 특성이 의도적인 기억 부호화에 미치는 영향을 알아보고자 사건-유관 fMRI 연구를 수행하였다. 부호화단계에서 구체어, 대상 그림, 그리고 추상패턴을 의도적으로 학습하도록 요구하면서 fMRI 영상을 구하였고 인출단계에서 재인판단검사를 실시하였다. 재인 수행 결과 그림, 구체어, 추상패턴의 순으로 기억 수행이 우수하였다. 상측 전두회(BA 6), 상측 두정피질(BA 7), 그리고 내측두피질의 세 ROI에서 신호변화율을 분석한 결과, 상측 전두회에서는 추상패턴만이 좌반구보다 우반구에서 더 큰 활성화를 보였고 구체어와 그림은 반구 차이를 보이지 않았으며 추상패턴은 우반구에서 구체어와 그림보다 더 큰 활성화를 보였다. 상측 두정피질에서는 세 학습자료 모두 반구 차이를 보이지 않았고 추상패턴은 구체어와 그림보다 더 큰 활성화를 보였으며, 내측두피질에서는 세 학습자료간 차이와 반구차이가 모두 관찰되지 않았다. 학습자료간 대비를 수행한 결과, 추상패턴은 신호변화율 분석 결과와 유사하게 상측 전두회에서는 구체어와 그림에 비해 우반구에서 더 큰 활성화를 보였고 상측 두정피질에서는 구체어에 비해 양반구에서, 그림에 비해 우반구에서 더 큰 활성화를 보였다. 그러나 내측두피질에서는 학습자료간 대비 결과가 신호변화율 분석 결과와 상이하였는데, 구체어와 그림은 모두 추상패턴에 비해 양반구에서, 그림은 구체어에 비해 우반구에서 더 큰 활성화를 보였다. 전반적으로 구체어와 그림의 활성화패턴은 서로 유사하였으며 추상패턴과는 공통적으로 상이한 활성화패턴을 보였는데, 이러한 결과는 자료-특수 편측성 가설보다는 부호-특수 편측성 가설을 지지하는 것으로서, 기억 부호화의 신경기전이 학습자료의 심상화와 언어화 가능성에 따라 달라진다는 것을 시사하였다.

저전력 비동기식 시스템 설계를 위한 혼합형 dual-rail data encoding 방식 제안 및 검증

지화준(Huajun Chi),김상만(Sangman Kim),박주성(Jusung Park) 대한전자공학회 2014 전자공학회논문지 Vol.51 No.7

본 논문에서는 dual-rail data encoding방식을 적용하여 비동기식시스템을 설계할 때, 신호천이를 줄이고 소비전력을 줄이기위하여 4-phase handshaking 프로토콜과 2-phase handshaking 프로토콜을 혼합한 dual-rail data encoding방식을 제안한다. 기존의 dual-rail data encoding 4-phase handshaking 프로토콜은 space state가 존재함으로 말미암아 신호 천이가 많이 발생하게 되고 많은 전력소비를 발생한다. 이론적으로 dual-rail data encoding 2-phase handshaking 프로토콜은 dual-rail data encoding 4-phase handshaking 프로토콜보다 빠르고 신호천이도 적지만 표준 라이브러리를 사용하여 설계할 수 없다. 제안하는 혼합형 dual-rail data encoding 방식의 성능을 평가하기 위하여 Adder블록, Multiplier블록, Latch를 포함한 benchmark회로를 설계를 설계하였다. Benchmark회로를 이용하여 시뮬레이션해본 결과, 제안하는 혼합형 dual-rail data encoding방식은 기존의 dual-rail data encoding 4-phase handshaking 프로토콜에 비해 35%이상 전력소비가 감소되는 결과를 얻었다. In this paper, we proposed new dual-rail data encoding that mixed 4-phase handshaking protocol and 2-phase handshaking protocol for asynchronous system design to reduce signal activities and power consumption. The dual-rail data encoding 4-phase handshaking protocol should leat to much signal activities and power consumption by return to space state. Ideally, the dual-rail data encoding 2-phase handshaking protocol should lead to faster circuits and lower power consumption than the dual-rail 4-phase handshaking protocol, but can not designed using standard library. We use a benchmark circuit that contains a multiplier block, an adder block, and latches to evaluate the proposed dual-rail data encoding. The benchmark circuit using the proposed dual-rail data encoding shows an over 35% reduction in power consumption with 4-phase dual-rail data encoding.

성경 히브리어 내러티브에서 오버인코딩 (Overencoding)에 대한 소고 - 사무엘상 1-7장을 중심으로 -

김성광(Seong-Kwang Kim) 한국구약학회 2019 구약논단 Vol.25 No.4

In the Biblical Hebrew narrative, the names of the participants can be generally referred by pronouns or verbs after they have first been introduced. However, we often see instances where characters are given semantically redundant information, such as renaming them. This paper tried to find out the author’s intention through a study on such redundant information. For this purpose, I introduced Runge’s recent study which is the most comprehensive and systematic study about participant reference encoding. Then, I investigated the function of over-encoding in 1 Samuel 1-7. I defined the default encoding value in the narrative, based on the results of cognitive linguistic studies. The use of redundant full noun phrase, where minimal encoding is expected, is to mark the over-encoding of participants. The use of redundant full noun phrase by participants is recognized by the reader as discontinuity. The functions of over-encoding are (1) marker of the beginning or ending of a unit, (2) highlighting action/event, (3) emphasizing the contents of speeches, (4) marker of countering moves, (5) marker of the significance of surprising or unexpected events, (6) marker of change in action pattern, (7) marker of focus shift, (8) marker of climax, (9) marker of contrast. All the other functions except marker of the beginning or ending of a unit are to highlight an element of the text. And they are recognized as pragmatic prominence by the reader through over-encoding intended by the author. That is, an element is highlighted by over-encoding, and more prominence is given to that element, as compared to other elements in the text. Over-encoding is a device, and prominence is the purpose for the use of over-encoding. As for the author, over-encoding is a device used for the production of the text, while, as for the reader, prominence is the result of recognition after the processing of the text. The significance of this study may be said in two. First, I synthesized scholars’ discussions so far on participant reference encoding in the Biblical Hebrew narrative as much as possible. As a result, the functions of over-encoding are divided in two: (1) marker of the beginning or ending of a unit, (2) highlighting an element. Second, I connected the highlighting function of over-encoding with prominence dealt with in pragmatics. For the reader, the over-encoded part of the text is recognized as more prominent than other parts. To find out over-encoding intended by the author will help the reader to interpret the text.

Evaluation of Spatial-encoding-related Geometric Distortion in Magnetic Resonance Imaging

홍철표,이동훈,이만우,우영근 한국물리학회 2012 THE JOURNAL OF THE KOREAN PHYSICAL SOCIETY Vol.61 No.12

This study demonstrated the characteristics of spatial encoding-related geometric distortion in magnetic resonance imaging (MRI), using a phantom with known physical dimensions to detect geometric distortion in MR scans. The amount of distortion was calculated as the difference between the physical coordinates of control points in the phantom and those of the corresponding points in the distorted MR image of the phantom. The phase and frequency encoding directions were swapped to acquire a phantom image for movement of a patient table 120쟭m to the right and left, allowing comprehensive distortion mapping over the isocentric plane and the entire field of view (FOV) along the spatial encoding. The geometric distortion of phase encoding directions over the entire FOV was small compared to the distortion of the frequency encoding directions. The maximum absolute deviations were 28.00 mm and 20.00 mm along the frequency and phase encoding directions over the entire FOV, respectively. The mean absolute deviations along the frequency and phase encoding directions were 2.85 mm and 1.97 mm, respectively. Although geometric distortion along the phase encoding axis near the isocenter was small, the distortion increased slightly toward the peripheral region. The distortion of the phase encoding direction in the peripheral region can be severely affected by imaging gradient nonlinearity.

교육수준에 따른 알츠하이머형 치매 환자의인지 부호화 사용의 차이: 언어적, 시공간적,동작적 부호화를 중심으로

곽경현,현명호,유정헌 한국임상심리학회 2006 Korean Journal of Clinical Psychology Vol.25 No.2

The primary purpose of this study was to investigated the differences of cognitive encoding among DAT(Dementia of Alzheimer's Type) patients' according to their educational level. In order to achieve this goal, subjects were presented categorizable list of words or objects under three different encoding condition; verbal, visual-spacial, motoric encoding. As a result of this research, the DAT patients have less memory perform than normal old people. And people who had lower level of education have less memory performance than who had higher level of education. Moreover, there was no difference between normal group and DAT group using memory ability in motoric encoding method according to the level of educational. But, in terms of verbal encoding and visual-spatial encoding methods, people who had more education have more memory ability. The present results suggested to the possibility of different education effect on the encoding methods.

TPM을 도입한 프로그램 카운터 인코딩의 구현

박수현(Soohyun Park),표창우(Changwoo Pyo),김선일(Sunil Kim),이경호(Gyungho Lee) 한국정보과학회 2011 정보과학회 컴퓨팅의 실제 논문지 Vol.17 No.1

프로그램 카운터 인코딩은 스택 스매슁을 비롯한 아직 알려지지 않은 코드 포인터에 대한 공격까지도 막을 수 있는 방법으로서, 코드 포인터 값을 저장할 때에는 암호화 하고, 프로그램 카운터 값이 프로그램 카운터로 적재되기 직전에는 복호화 하는 방식이다. 프로그램 카운터 인코딩의 신뢰도는 암호화 할 때 사용하는 키 값과 키 저장 위치의 안전성에 의해 결정된다. 본 논문에서는 프로그램 카운터 인코딩의 키의 신뢰성을 강화하기 위해 신뢰성 컴퓨팅 그룹(Trusted Computing Group)의 TPM(Trusted Platform Module)을 도입한 구현 사례를 제시하고, 평가하였다. 스택과 힙에서 버퍼 오버플로우나 포맷스트링을 이용한 반환 주소에 대한 모의 공격은 모두 방어할 수 있음이 검증되었다. 본 논문에서 구현한 방법은 SPEC CPU2006을 사용한 성능 시험에서는 3% 정도의 과부하를 보였다. 구현 방법은 TPM의 성능 부담과 보안 신뢰성을 고려하여 제안하였다. Program counter encoding is a method that can prevent code pointer attacks as yet unknown including the stack smashing. Program counter encoding enforces encoding code pointer values when they are stored and decoding when they are loaded into the program counter. Safety of the key values and storage used for encoding code pointers determines the trustworthiness of program counter encoding. This paper presents and evaluates an implementation of program counter encoding with TPM (Trusted Platform Module) to reinforce the trustworthiness of the keys for program counter encoding. Simulated attacks against return addresses on stack and heap area confirmed defensive capability of program counter encoding. Performance of our implementation shows 3% overhead in SPEC CPU2006. The suggested implementation is based on the considerations of both performance and trustworthiness.

Encoding Selection for a Class of Fitness Functions based on Locus Interdependency

Hongqiang Mo,Zhong Li,박진배,주영훈,Qiliang Du 제어·로봇·시스템학회 2015 International Journal of Control, Automation, and Vol.13 No.5

The feasible solutions are usually represented with a binary encoding when a genetic algorithm is applied to a continuous optimization problem. However, it has been remarked that a binary encoding was not always the best choice, and it was suggested to use a base-m encoding for a class of fitness functions linearly combined of sine functions whose frequencies were exponential to a positive integer m. In this paper, this suggestion is explained based on locus interdependency. It is shown that, for these fitness functions, the Euclidean distances from a considerable part of the highly fit strings to the objective strings are negative powers of m. Thus, the Hamming distances from the highly fit strings to the objective strings when the feasible solutions of these fitness functions are represented with a basem encoding are much smaller than those when the fitness functions are expressed with an encoding of another cardinality. And as a result, locus interdependency of the former is much lower than that of the latter, which indicates that the fitness functions are likely to be much easier when expressed with the former encoding. The suggestion is then tested on a number of fitness functions randomly generated, in which encodings with different bases are compared according to locus interdependency and optimization performance. The results of the test substantiate the suggestion.

Graphical and SERS dual-modal identifier for encoding OBOC library

Kang, Homan,Koh, Yul,Jeong, Sinyoung,Jeong, Chulhwan,Cha, Myeong Geun,Oh, Min-Hye,Yang, Jin-Kyoung,Lee, Hyunmi,Jeong, Dae Hong,Jun, Bong-Hyun,Chang, Hyejin,Kim, Yong-Kweon,Lee, Yoon-Sik Elsevier 2020 Sensors and actuators. B, Chemical Vol.303 No.-

<P><B>Abstract</B></P> <P>‘Split-and-mix’ synthetic methods are widely utilized to prepare extensively large numbers of combinatorial libraries. For this, a simple and reliable encoding/decoding method is required to identify a lead compound, especially when it comes to combinatorial peptide/peptoid libraries. Here, we report a simple, efficient, and reliable on-bead peptoid ligand identification method using dual-modal identifiers (DM-IDs), which have a graphical pattern and a Raman signal. The DM-ID was designed to encode the sequence and the type of peptoid side chain by the number of holes and the surface-enhanced Raman scattering (SERS) signals, respectively. Using this dual encoding strategy, 5832 (18<SUP>3</SUP>)-individual peptides/peptoids can be encoded with 3 holes and 18 Raman label compounds. In addition, the method can be easily extended to encode more than one hundred billion ligands by expanding to 9 patterns with 3 holes in binary code. After pentapeptoids were encoded with DM-IDs during solid phase synthesis, the peptoid’s affinity towards streptavidin is successfully evaluated, and the on-bead peptoid sequence is determined by the graphical patterns and the SERS signatures of the DM-IDs. We believe that this encoding method using DM-IDs is a promising tool to construct one-bead-one-compound (OBOC) combinatorial libraries for drug discovery.</P> <P><B>Highlights</B></P> <P> <UL> <LI> On-bead peptoid ligand identification method using dual-modal identifiers (DM-IDs) is demonstrated. </LI> <LI> The DM-ID with graphical (hole structure) and optical (SERS spectrum) codes is successfully fabricated. </LI> <LI> The DM-ID is designed to dictate the sequence of peptoid ligand and the type of side chain. </LI> <LI> More than one hundred billion (18<SUP>9</SUP>)-individual ligands can be encoded with 9-hole patterns and 18 Raman label compounds. </LI> </UL> </P>

지각적 부호화와 의미적 부호화가 주관적 인지저하의 자유회상과 재인회상에 미치는 효과

김해윤,강연욱,유경호,이병철 한국임상심리학회 2018 Korean Journal of Clinical Psychology Vol.37 No.3

The present study was conducted to examine the differential effects of perceptual and semantic encodings on recall and recognition memory in elderly people with Subjective Cognitive Decline (SCD) and to compare them with those in the normal elderly (NE) and amnestic Mild Cognitive Impairment (amMCI). Computerized perceptual (number of letters) and semantic (man-made/natural) verbal encoding tests were administered. After a 20-minute delay, free-recall and recognition tests for the encoded target words were performed. At the comparison of the difference (semantic encoding-perceptual encoding) of memory performances between perceptual and semantic encodings, the NE showed a greater difference than the amMCI, but the SCD did not show any differences with either NE or amMCI. Although the amMCI showed poorer performance on both recall and recognition tests than the NE and SCD, the SCD did not show any differences in both tests with NE. At the comparison of the difference (recognition-delayed free recall) of the memory performance between the delayed free-recall and recognition, the NE showed a greater difference than the amMCI, but the SCD did not show the difference with either group. These results showed that the SCD already had subtle problems in semantic encoding and memory retrieval but with less degree than amMCI. They suggest that the SCD is an intermediate stage between the NE and amMCI.

Modeling and Simulation of New Encoding Schemes for High-Speed UHF RFID Communication

모상현,배지훈,박찬원,방효찬,박형철 한국전자통신연구원 2015 ETRI Journal Vol.37 No.2

In this paper, we present novel high-speed transmission schemes for high-speed ultra-high frequency (UHF) radio-frequency identification communication. For high-speed communication, tags communicate with a reader using a high-speed Miller (HS-Miller) encoding and multiple antennas, and a reader communicates with tags using extended pulse-interval encoding (E-PIE). E-PIE can provide up to a two-fold faster data rate than conventional pulse-interval encoding. Using HS-Miller encoding and orthogonal multiplexing techniques, tags can achieve a two- to three-fold faster data rate than Miller encoding without degrading the demodulation performance at a reader. To verify the proposed transmission scheme, the MATLAB/Simulink model for high-speed backscatter based on an HS-Miller modulated subcarrier has been designed and simulated. The simulation results show that the proposed transmission scheme can achieve more than a 3 dB higher BER performance in comparison to a Miller modulated subcarrier.