http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Single Image Depth Estimation With Integration of Parametric Learning and Non-Parametric Sampling
Jung, Hyungjoo,Sohn, Kwanghoon Korea Multimedia Society 2016 멀티미디어학회논문지 Vol.19 No.9
Understanding 3D structure of scenes is of a great interest in various vision-related tasks. In this paper, we present a unified approach for estimating depth from a single monocular image. The key idea of our approach is to take advantages both of parametric learning and non-parametric sampling method. Using a parametric convolutional network, our approach learns the relation of various monocular cues, which make a coarse global prediction. We also leverage the local prediction to refine the global prediction. It is practically estimated in a non-parametric framework. The integration of local and global predictions is accomplished by concatenating the feature maps of the global prediction with those from local ones. Experimental results demonstrate that the proposed method outperforms state-of-the-art methods both qualitatively and quantitatively.
지역혁신 활성화를 위한 도시기반 혁신정책의 전략과 방향
김형주(Hyungjoo Kim),정미애(Mi-Ae Jung),최해옥(Hae Ok Choi),임영훈(Young-Hun Lim),고병옥(Byeung-Ok Ko) 과학기술정책연구원 2017 정책연구 Vol.- No.-
Cities are becoming a center of innovations in the age of Digital Transformation. Urban areas are recognized as a laboratory and platform of innovations in digital spaces as well as in physical ones. Cities have an advantage over an isolated campus-type rural areas as a place of innovations since the high density and diversity urban settings create contribute to network-based innovations. Urban areas, also, are the location of concentrated data creation and utilization. Data sciences and Information Technologies are recently solving problems in urban areas and improving public services. This study examines the significance of urban innovation in the age of Digital Transformation and understands new opportunities of place-based innovation policies. The first part of the research focuses on ‘innovation districts’ from the perspective of physical spaces; the second part investigates trends and issues of data-driven urban innovation in digital spaces. The study provides policy directions and strategies of for urban innovation.
Deep Monocular Depth Estimation via Integration of Global and Local Predictions
Kim, Youngjung,Jung, Hyungjoo,Min, Dongbo,Sohn, Kwanghoon IEEE 2018 IEEE TRANSACTIONS ON IMAGE PROCESSING - Vol.27 No.8
<P>Recent works on machine learning have greatly advanced the accuracy of single image depth estimation. However, the resulting depth images are still over-smoothed and perceptually unsatisfying. This paper casts depth prediction from single image as a parametric learning problem. Specifically, we propose a deep variational model that effectively integrates heterogeneous predictions from two convolutional neural networks (CNNs), named global and local networks. They have contrasting network architecture and are designed to capture the depth information with complementary attributes. These intermediate outputs are then combined in the integration network based on the variational framework. By unrolling the optimization steps of Split Bregman iterations in the integration network, our model can be trained in an end-to-end manner. This enables one to simultaneously learn an efficient parameterization of the CNNs and hyper-parameter in the variational method. Finally, we offer a new data set of 0.22 million RGB-D images captured by Microsoft Kinect v2. Our model generates realistic and discontinuity-preserving depth prediction without involving any low-level segmentation or superpixels. Intensive experiments demonstrate the superiority of the proposed method in a range of RGB-D benchmarks, including both indoor and outdoor scenarios.</P>
Astrocyte elevated gene-1 overexpression in hepatocellular carcinoma
Hae Il Jung,Taesung Ahn,Sang Ho Bae,Jun Chul Chung,Hyungjoo Kim,Susie Chin,Dongjun Jeong,Hyon Doek Cho,Moon Soo Lee,Hyung Chul Kim,Chang Ho Kim,Moo-Jun Baek 대한외과학회 2015 Annals of Surgical Treatment and Research(ASRT) Vol.88 No.2
Purpose: Astrocyte elevated gene-1 (AEG-1) plays important roles in tumorigenesis such as proliferation, invasion, metastasis, angiogenesis, and chemoresistance. We examined the expression of AEG-1 in patients with hepatocellular carcinoma (HCC). Methods: Eighty-five samples were collected from patients with HCC who underwent surgery and were histopathologically confirmed to have HCC. Two independent pathologists, experienced in evaluating immunohistochemistry and blinded to the clinical outcomes of the patients, reviewed all samples. They determined AEG-1 expression semiquantitatively by assessing the percentage of positively stained immunoreactive cells and staining intensity. Clinicopathological data were analyzed in association with prognosis. Results: The association was estimated by univariate and multivariate analyses with Cox regression. Tumor size (hazard ratio [HR], 2.285; 95% confidence interval [CI], 1.175?4.447; P = 0.015), microvascular invasion (HR, 6.754; 95% CI, 1.631?27.965; P = 0.008), and AEG-1 expression (HR, 4.756; 95% CI, 1.697?13.329; P = 0.003) were independent prognostic factors for overall survival. Those for disease-free survival rate were tumor size (HR, 2.245; 95% CI, 1.282?3.933; P = 0.005) and AEG-1 expression (HR, 1.916; 95% CI, 1.035?3.545; P = 0.038). The cumulative 5-year survival and recurrence rates were 89.2% and 50.0% in the low-expressing group and 24.5% and 82.4% in the high-expressing group, respectively. Conclusion: The results suggest that AEG-1 overexpression could serve as a valuable prognostic marker in patients with HCC.