http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
김명준(MyungJun Kim),이영렬(Yung-Lyul Lee) 한국방송·미디어공학회 2017 방송공학회논문지 Vol.22 No.3
HEVC 표준은 정수 화소로 표현된 신호에 DCT-II를 기반으로 하는 보간 필터를 사용하여 부화소 신호를 생성한다. 이러한 방법으로 생성된 신호는 움직임 보상 및 예측의 성능 향상을 가져온다. HEVC 표준은 부화소를 생성하기 위해서 길이가 다른 각각의 DCT보간 필터를 사용하고 있다. 1/2-화소를 생성할 땐, 필터의 길이가 8인 DCT 기반 보간 필터를 사용하며, 1/4-화소와 3/4-화소의 경우에는 필터의 길이가 7인 DCT 기반 보간 필터를 사용한다. 본 논문에서는 DST-VII을 기반으로 하는 보간 필터를 제안하여, 움직임 보상 및 예측의 성능 향상을 가져온다. 본 논문에서 제안하는 방법은 HEVC 표준보다 BD-rate가 Random Access와 Low Delay B configurations에서 각각 0.45%와 0.5%의 성능 향상을 가져온다. High Efficiency Video Coding (HEVC) adopted the Discrete Cosine Transform-II (DCT-II) based interpolation filter to improve coding efficiency in motion compensation and estimation. In HEVC, the interpolation filters based on the DCT-II are composed of 8-point for half-pixel and 7-point for 1/4-pixel and 3/4-pixel. In this paper, a DST-VII based interpolation filter is used improve motion compensation and estimation. The experimental results which applied the DST-VII interpolation filter are presented. They show the 0.45% of average bitrate reduction in Random Access configuration and 0.5% of average bitrate reduction in Low Delay B configuration, respectively.
Semi-Supervised Learning for Stratified Networks
Myungjun Kim(김명준),Yonghyun Nam(남용현),Sangkuk Lee(이상국),Hyunjung Shin(신현정) 한국경영과학회 2016 한국경영과학회 학술대회논문집 Vol.2016 No.4
This research is concerned with developing a semi-supervised learning algorithm for stratified networks. In stratified networks, labels in one stratum can benefit predictions in other strata through inter-stratum connections so dealing with inter-stratum connections is important. Technically, the problem of non-squareness and sparseness involved in matrix inversion for interstratum connections must be solved. In order to verify the validity of the algorithm, it was applied on disease-symptom network structure to predict cooccurrence of two diseases.
Semi-Supervised Learning for Stratified Networks
Myungjun Kim(김명준),Yonghyun Nam(남용현),Sangkuk Lee(이상국),Hyunjung Shin(신현정) 대한산업공학회 2016 대한산업공학회 춘계학술대회논문집 Vol.2016 No.4
This research is concerned with developing a semi-supervised learning algorithm for stratified networks. In stratified networks, labels in one stratum can benefit predictions in other strata through inter-stratum connections so dealing with inter-stratum connections is important. Technically, the problem of non-squareness and sparseness involved in matrix inversion for interstratum connections must be solved. In order to verify the validity of the algorithm, it was applied on disease-symptom network structure to predict cooccurrence of two diseases.