VLIW 기반 고성능 DSP에서의 SAD 알고리즘 최적화 스케쥴링

유희재(Hui-Jse Yu),정선태(Sun-Tae Chung),정수환(Souhwan Jung) 한국콘텐츠학회 2007 한국콘텐츠학회논문지 Vol.7 No.12

SAD(Sum of Absolute Difference) 알고리즘은 동영상 인코더에서 가장 많은 시간이 소용되는 것으로 잘 알려진 움직임 추정에서 가장 자주 계산이 수행되는 알고리즘으로, 동영상 인코딩 수행시간을 줄이기 위해서 우선적으로 최적화 구현되어야 하는 알고리즘이다. 본 논문에서는 VLIW 기반 고성능 DSP 프로세서에서의 조건 분기를 갖는 SAD 알고리즘의 최적 스케줄링 구현 방법을 제안한다. 제안 방법은 먼저 조건분기를 갖는 중첩 루프를 VLIW 구조가 제공하는 ILP(Instruction Level Parallelism) 능력을 잘 활용할 수 있도록 충분한 크기의 루프 몸체를 가지며 또한 빨리 루프를 탈출 할 수 있는 조건 분기를 갖는 단일 루프로 변환한 후에, 모듈로 스케줄링 기법을 적용하여 VLIW 기반 프로세서에서 최적화 스케줄링 구현을 한다. 제안된 구현 방안을 TMS320C6713에서 구현하고, 코드 크기 및 수행 시간에 대한 성능 분석을 하였다. 구현된 최적화 SAD 루틴은 코드 크기도 크지 않아 임베디드 응용에 적합하며, 이 SAD 구현을 사용한 H.263 인코더가 그렇지 않은 H.263 인코더보다 훨씬 좋은 성능을 보임을 실험을 통해 확인하였다. SAD (Sum of Absolute Difference) algorithm is the most frequently executing routine in motion estimation, which is the most demanding process in motion picture encoding. To enhance the performance of motion picture encoding on a VLIW processor, an optimal implementation of SAD algorithm on VLIW processor should be accomplished. In this paper, we propose an implementation of optimal scheduling of SAD algorithm with conditional branch on a VLIW-based high performance DSP. We first transform the nested loop with conditional branch of SAD algorithm into a single loop with conditional branch which has a large enough loop body to utilize fully the ILP capability of VLIW DSP and has a conditional branch to make the escape from loop to be achieved as soon as possible. And then we apply a modulo scheduling technique to the tranformed single loop. We test the proposed implementation on TMS320C6713, and analyze the code size and performance with respect to processing time. Through experiments, it is shown that the SAD implementation proposed in this paper has small code size appropriate for embedded applications, and the H.263 encoder with the proposed SAD implementation performs better than other H.263 encoder with other SAD implementations.

남성과 여성을 나타내는 동물 은유의 비판적 분석

유희재(Yu, Huijae) 담화·인지언어학회 2017 담화와 인지 Vol.24 No.4

The goal of this analysis is to figure out why a man is a wolf and a woman is a fox. Specifically, it aims (1) to establish a list of metaphors that portray men and women as animals and (2) to critically uncover ideologies embedded in the metaphors. In the data from the TV show 〈Witch Hunt〉, men and women are mapped into different animals in the metaphors. Men are portrayed as a general animal as a statue, and other specific animals such as a chimpanzee, a horse, a tiger, a wolf or a worker bee. On the other hand, women are not mapped into a general animal, but into specific animals such as a bear, a fox, a queen bee, a snake, a cat or a dog. With this metaphor, it is justified that men should follow their ‘instinct’ as wild animals without any responsibility as a human being. Also, women are not only portrayed as lower animals than men’s but also animals which are either merely passive subjects to a possession or evil beings.

말뭉치언어학 관점에서 신문기사 범죄사건 명명 방식 분석

유희재(Yu, Huijae),유성희(Yoo, Seonghee),김진웅(Kim, Jinung) 담화·인지언어학회 2016 담화와 인지 Vol.23 No.2

This study undertakes a corpus-based methodology to analyze the naming process of crime cases in the headlines of Korean newspapers. The most used piece of information in the name of crime cases was information about the crime itself, often realized as ‘murder’ or ‘rape’ cases. This information was often used together with information about the location of the crime, which was found to be the second most used piece of information. This tendency can be explained by the optimization of conveying information - more general information about the crime type together with more specific information about its location can refer to a specific instance of crime most effectively. It was found in the data that in the naming of crime cases, information about the perpetrators of the crime was used much less frequently than information about the victims of those crimes, with later analysis revealing that only in unusual cases were details of criminals given. Information about the victims on the other hand was used widely, especially when the victims were female and of a young age. Following these findings, we underline that the ‘ideal’ victims of crime for Korean newspapers to report are young women, and that the media tend to emphasize the victims of crime rather than the criminals.

옥타데실트리클로로실란 코팅에 의한 실리콘 표면 특성 변화

유희재(Yu Hui Jae),김수경(Kim Su Gyeong),김진홍(Kim Jin Heung),강호종(Kang Ho Jong) 한국고분자학회 2003 폴리머 Vol.27 No.6

다문화가정 미취학 아동 말하기,듣기 이중언어 교육 프로그램 개발을 위한 기초 연구

유희재 ( Hui Jae Yu ),최윤영 ( Yoon Young Choi ),카제노바 아셀 ( Kadenova Asel ),강승혜 ( Seung Hae Kang ) 연세대학교 언어연구교육원 한국어학당 2016 외국어로서의 한국어교육 Vol.44 No.-

This study includes a basic research for a program design aiming at 3- to 5-year-old preschoolers of multi-cultural families under bilingual environment. For the program design, this study conducts a multilateral needs analysis including a literature review, analyses of previous programs, and interviews with the parents and an employee from a related institution, and classroom observations as well. With the interview with multi-cultural families we figured out that their children have more problems with the language of their foreign mothers rather than Korean. Also the parents responded that they would like the program to teach children similar things as what native Korean children at that age are taught which can be found from pre-existing Korean kindergarten programs. The employee from the institution suggested that a bilingual program for children would work the best when a teacher directly teaches the children. At the later part of this paper we suggest a curriculum focusing on listening and speaking ability based on the result of the needs analysis.

<메르스>에 대한 은유와 이데올로기적 함축

전혜영(Hyeyoung Jeon),유희재(Huijae Yu) 한국어학회 2016 한국어학 Vol.72 No.-

This study has two main purposes: to establish a list of source domains in the metaphors for Middle East respiratory syndrome (MERS) and to uncover ideological meanings embedded in them in Korean news reports from KBS and JTBC. The first part of this study presents metaphors such as [MERS IS WAR], [MERS IS WAVE], [MERS IS A LIVING THING], and [MERS IS A THING], which were found in the data. The latter part of this study deals with how the two broadcasting companies use these metaphors differently according to their ideologies. In the metaphor of [MERS IS WAR], KBS tends to show less of the agents who controls the war since the war against MERS has failed which casts responsibility to the controlling agents, the government and big hospitals. In this, KBS tries to present less of the information of the responsible agents that presented in JTBC. Through the metaphor of [MERS IS WAVE], KBS presents the aftermath of MERS as something not serious. Compared to JTBC, KBS tends to suggest that the aftermath of MERS is predominantly an economic effects by metaphorically suggesting that predominantly the economic sector got hit by MERS.

대화형 인공지능의 윤리적 언어 표현을 위한 기초 연구

조태린(Cho, Tae-Rin),김신각(Kim, Shin-Gak),유희재(Yu, Hui-Jae),김예지(Kim, Ye-Ji),이주희(Lee, Ju-Hee) 한국어문학회 2018 語文學 Vol.0 No.140

The purpose of this study is to set up the concept and classification standard of unethical linguistic expressions(ULEs) as word unit that can appear in the machine learning of interactive artificial intelligence systems and humans conversation. The study also examined the characteristics of each type, based on examples. According to the results of the study, ULEs as word unit can be divided into three categories: ‘swear words’, ‘vulgar words’, and ‘derogatory slangs’. These three types are common in that they are all regarded as ULEs in any variance of emergence. However, they differ in two attributes, aggression and degradation. This study also suggests that there are two other types of ULEs: the ‘Context’ type, in which certain expressions may or may not be used unethically according to the context, and the ‘Content’ type, which can not be considered unethical linguistic expression in of itself, but its related content can be unethical linguistic expression overall. Therefore, this study identifies that the word-based unethical linguistic expression is classified into five types such as ‘swear words’, ‘vulgar words’, ‘derogatory slang’, ‘context’, and ‘content’, and each of these characteristics and examples were reviewed extensively.