http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
김호준 한국뇌학회 2003 한국뇌학회지 Vol.3 No.1
본 연구에서는 보다 개선된 영상분할을 얻기 위한 방법으로서, 에지정보와 영역정보가 상호 결합하는 모듈구조의 신경망 기법을 개발하였다. 제안된 모델에서는 가보(Gabor) 필터와 자기구조화특징지도(SOFM) 모델이 초기화 단계를 위해 사용되며, 셀룰러 신경망 및 CSNN이 반복단계의 영상분할과정을 위해 서로 결합하여 동작한다. We have been developed a tracking model to improve the recognition performance, We designed a top-down based face detection model that receives the input data from the bottom-up visual attention network, In order to detect a face effectively, the symmetry feature is applied to visual attention model, Also, we designed a visual tracking system that combines the visual attention and the eye movement.
Bottom up approach of artificial exosomes
김호준,박성욱,정영도,이관희,이효진 한국공업화학회 2019 한국공업화학회 연구논문 초록집 Vol.2019 No.0
Exosomes are small (<150 nm) vesicles secreted by various cells in humn body. Recent advances of nanotechnology unveiled that exosomes contain important biomarkers including miRNAs and different types of proteins specific to parental cells. Hence, increased number of studies are reported therapeutic potential of exosomes as a stable and reliable biomarker for various diseases. Because of the innate complexity of exosomes, translational research is limited and one of the biggest challenge of exosome is heterogeneity. To tackle the challenge, we synthesized artificial exosomes with desired physicochemical properties through microfluidic mixer deisgn. Specifically, artificial exosomes with 40, 70, and 100 nm size and low OID (<0.1) are succssfully synthesized with the desired densities of memebrane proteins and membrane charge densities. We will discuss how the exosomes' size, charge density, and concentration of membrane proteins are correlated to biological functions.