243 Hits in 1.5 sec

Deep Iterative Frame Interpolation for Full-frame Video Stabilization [article]

Jinsoo Choi, In So Kweon
2019 arXiv   pre-print
Video stabilization is a fundamental and important technique for higher quality videos. Prior works have extensively explored video stabilization, but most of them involve cropping of the frame boundaries and introduce moderate levels of distortion. We present a novel deep approach to video stabilization which can generate video frames without cropping and low distortion. The proposed framework utilizes frame interpolation techniques to generate in between frames, leading to reduced inter-frame
more » ... jitter. Once applied in an iterative fashion, the stabilization effect becomes stronger. A major advantage is that our framework is end-to-end trainable in an unsupervised manner. In addition, our method is able to run in near real-time (15 fps). To the best of our knowledge, this is the first work to propose an unsupervised deep approach to full-frame video stabilization. We show the advantages of our method through quantitative and qualitative evaluations comparing to the state-of-the-art methods.
arXiv:1909.02641v1 fatcat:ejaza77lyzf3pax7mgv65gflfi

Space-Borne Monitoring of NOx Emissions from Cement Kilns in South Korea

Hyun Cheol Kim, Changhan Bae, Minah Bae, Okgil Kim, Byeong-Uk Kim, Chul Yoo, Jinsoo Park, Jinsoo Choi, Jae-bum Lee, Barry Lefer, Ariel Stein, Soontae Kim
2020 Atmosphere  
Nitrogen oxide (NOx) emissions from the South Korean cement industry are investigated with remote-sensing measurements, surface observations, and in situ aircraft measurements. In the Yeongwol, Danyang, and Jecheon regions of central South Korea, six closely located cement factories produce 31 million tons of cement annually. Their impact on the regional environment has been a public-policy issue, but their pollutants have not been continuously monitored nor have emissions inventories been
more » ... verified. Using a newly developed downscaling technique, remote-sensing analyses show that Ozone Monitoring Instrument (OMI) NO2 column densities over the cement kilns have more than twice the modeled concentrations, indicating that the kilns are one of the most dominant NOx emission point sources in South Korea. Observed NOx emissions are stronger in the spring, suggesting that these sources play an important role in the formation of surface ozone and secondary particulate matter. These emissions also slightly increased in recent years, even while most major South Korean cities posted a declining trend in NOx emissions. Photochemical models (during May to July 2015) demonstrate that emissions from the South Korean cement industry have significant environmental impacts, both on surface ozone (up to approximately 4 ppb) and PM2.5 (up to approximately 2 µg/m3).
doi:10.3390/atmos11080881 fatcat:2o4yqduytrgbvpro5xmiq3t3ge

Enhanced Single-Sided Ventilation with Overhang in Buildings

Jinsoo Park, Jung-Il Choi, Gwang Rhee
2016 Energies  
Enhancing the ventilation performance of energy-efficient buildings with single-sided openings is important because their ventilation performance is poor and strongly depends on the wind conditions. We considered an overhang as a potential building façade for improving the single-sided ventilation performance. We performed numerical simulations of three-dimensional unsteady turbulent flows over an idealized building with an overhang in order to investigate the effect of the overhang on the
more » ... lation performance. Parametric studies were systematically carried out where the overhang length, wind speed, and wind direction were varied. The numerical results showed that the overhang drastically enhanced the ventilation rate in the windward direction regardless of the wind speed. This is because, for windward cases, the overhang produces a vortex with strong flow separation near the tip of the overhang, which promotes a net airflow exchange at the entrance and increases the ventilation rate. However, the ventilation rates for the leeward and side cases are slightly decreased with the overhang. Using an overhang with single-sided ventilation greatly reduces the local mean age of air (LMA) in the windward direction but increases it in the leeward direction.
doi:10.3390/en9030122 fatcat:4rpnvrxrqnferei4ekrrggcukm

Radiopaque markers and hydrogram in feline alimentary lymphoma

Jihye Choi, Jinsoo Lee, Jaeyoung Jang, Heeyeon Choi, Jimin Seo, Minjung Lee, Hyunwook Kim
2012 Korean Journal of Veterinary Research  
Radiography, ultrasonography with hydrogram, and contrast studies using radiopaque markers were applied to evaluate alimentary lymphoma in two cats. The hydrogram facilitated the differentiation of pseudo-thickening from true wall thickening, and enabled an evaluation of wall layering and lymph nodes. In case 1, mechanical obstruction of the duodenum was confirmed with barium-impregnated polyethylene spheres (BIPS), a radiopaque marker; however, results obtained in case 2 were not as
more » ... We expect that hydrograms and BIPS can be used as valuable alternative methods to evaluate the gastrointestinal (GI) tract although further studies in cases involving GI tumors are needed.
doi:10.14405/kjvr.2012.52.2.147 fatcat:ycz7xdl3kzgtthbhhzwh2gr5ey

Human Attention Estimation for Natural Images: An Automatic Gaze Refinement Approach [article]

Jinsoo Choi, Tae-Hyun Oh, In So Kweon
2016 arXiv   pre-print
Photo retrieval by saliency JinsooFig. 1 : 1 Choi, Tae-Hyun Oh and In So Kweon (corresponding author) are with the School of Electrical Engineering, Korea Advanced Institute of Science and Technology (  ... 
arXiv:1601.02852v1 fatcat:ixzxb3u7vneffllcb3jjtl5mja

Aircraft Measurements of Physicochemical Evolution of Atmospheric Aerosols in Air Pollution Plumes over a Megacity and Suburban Areas

Taehyun Park, Yongjoo Choi, Jinsoo Choi, Junyoung Ahn, Jinsoo Park, Yonghwan Lee, Jihee Ban, Gyutae Park, Seokwon Kang, Kyunghoon Kim, Beom-Keun Seo, Jongho Kim (+4 others)
2020 Aerosol and Air Quality Research  
Choi et al. (2016) identified the sources of organic aerosol (OA) in PM 10 (particulate matter with a diameter of < 10 µm) in Seoul, Korea, in 2010 Korea, in -2011 , indicating that most OA was emitted  ... 
doi:10.4209/aaqr.2019.12.0649 fatcat:tkn3lxjny5bdrc2cahkucmhx4a

Disjoint Multi-task Learning between Heterogeneous Human-centric Tasks [article]

Dong-Jin Kim, Jinsoo Choi, Tae-Hyun Oh, Youngjin Yoon, In So Kweon
2018 arXiv   pre-print
Human behavior understanding is arguably one of the most important mid-level components in artificial intelligence. In order to efficiently make use of data, multi-task learning has been studied in diverse computer vision tasks including human behavior understanding. However, multi-task learning relies on task specific datasets and constructing such datasets can be cumbersome. It requires huge amounts of data, labeling efforts, statistical consideration etc. In this paper, we leverage existing
more » ... ingle-task datasets for human action classification and captioning data for efficient human behavior learning. Since the data in each dataset has respective heterogeneous annotations, traditional multi-task learning is not effective in this scenario. To this end, we propose a novel alternating directional optimization method to efficiently learn from the heterogeneous data. We demonstrate the effectiveness of our model and show performance improvements on both classification and sentence retrieval tasks in comparison to the models trained on each of the single-task datasets.
arXiv:1802.04962v1 fatcat:flqul2wjrbax7gndviagr5atjy

ACP++: Action Co-occurrence Priors for Human-Object Interaction Detection [article]

Dong-Jin Kim, Xiao Sun, Jinsoo Choi, Stephen Lin, In So Kweon
2021 arXiv   pre-print
A common problem in the task of human-object interaction (HOI) detection is that numerous HOI classes have only a small number of labeled examples, resulting in training sets with a long-tailed distribution. The lack of positive labels can lead to low classification accuracy for these classes. Towards addressing this issue, we observe that there exist natural correlations and anti-correlations among human-object interactions. In this paper, we model the correlations as action co-occurrence
more » ... ces and present techniques to learn these priors and leverage them for more effective training, especially on rare classes. The efficacy of our approach is demonstrated experimentally, where the performance of our approach consistently improves over the state-of-the-art methods on both of the two leading HOI detection benchmark datasets, HICO-Det and V-COCO.
arXiv:2109.04047v1 fatcat:7xxsvg2lcned3jtgvxiyz6ehw4

Energy gap–refractive index relations in semiconductors – An overview

N.M. Ravindra, Preethi Ganapathy, Jinsoo Choi
2007 Infrared physics & technology  
An overview of the understanding of correlations between energy gap and refractive index of semiconductors is presented here. The atomic approach of Moss, the nearly free electron model of Penn, the oscillator concept of Wemple and the optical polarizability approach of Finkenrath are considered in this study. The Ravindra relation is discussed in the context of alternate approaches that have been presented in the literature. Case studies of applications of these relations to infrared materials and wide band gap semiconductors are presented.
doi:10.1016/j.infrared.2006.04.001 fatcat:24uwd5rrbnax3glqpafzhvq36u

Quantum key distribution between two groups using secret sharing [article]

Sora Choi, Jinsoo Kim, Dong Pyo Chi
2004 arXiv   pre-print
In this paper, we investigate properties of some multi-particle entangled states and, from the properties applying the secret sharing present a new type of quantum key distribution protocols as generalization of quantum key distribution between two persons. In the protocols each group can retrieve the secure key string, only if all members in each group should cooperate with one another. We also show that the protocols are secure against an external eavesdropper using the intercept/resend strategy.
arXiv:quant-ph/0306067v2 fatcat:dubghlobpba3joezgqmiu6lylq

Genomic Analysis Using Bayesian Methods under Different Genotyping Platforms in Korean Duroc Pigs

Jungjae Lee, Yongmin Kim, Eunseok Cho, Kyuho Cho, Soojin Sa, Youngsin Kim, Jungwoo Choi, Jinsoo Kim, Junki Hong, Taejeong Choi
2020 Animals  
Genomic evaluation has been widely applied to several species using commercial single nucleotide polymorphism (SNP) genotyping platforms. This study investigated the informative genomic regions and the efficiency of genomic prediction by using two Bayesian approaches (BayesB and BayesC) under two moderate-density SNP genotyping panels in Korean Duroc pigs. Growth and production records of 1026 individuals were genotyped using two medium-density, SNP genotyping platforms: Illumina60K and
more » ... 80K. These platforms consisted of 61,565 and 68,528 SNP markers, respectively. The deregressed estimated breeding values (DEBVs) derived from estimated breeding values (EBVs) and their reliabilities were taken as response variables. Two Bayesian approaches were implemented to perform the genome-wide association study (GWAS) and genomic prediction. Multiple significant regions for days to 90 kg (DAYS), lean muscle area (LMA), and lean percent (PCL) were detected. The most significant SNP marker, located near the MC4R gene, was detected using GeneSeek80K. Accuracy of genomic predictions was higher using the GeneSeek80K SNP panel for DAYS (Δ2%) and LMA (Δ2–3%) with two response variables, with no gains in accuracy by the Bayesian approaches in four growth and production-related traits. Genomic prediction is best derived from DEBVs including parental information as a response variable between two DEBVs regardless of the genotyping platform and the Bayesian method for genomic prediction accuracy in Korean Duroc pig breeding.
doi:10.3390/ani10050752 pmid:32344859 pmcid:PMC7277155 fatcat:timxay4tjzce5dqqbohphqjyke

일터에서의 그릿에 대한 이해

JinSoo Choi, SuRan Lee, YoungWoo Sohn
2021 The Korean Journal of Industrial and Organizational Psychology  
통해 일 중독에 영향을 미친다는 선행연구 마지막으로 직업 지향 그릿 군집은 직업 (Choi, Permpongaree, Kim, Choi, & Sohn, 2020) 지향성이 높고, 경력 지향성, 소명 지향성이  ...  ., 1997) 증하였다 (Choi, Sohn, & Lee, 2020; Duckworth et al., 2007; Lee & Sohn, 2013; Lim, 2017) (Wrzesniewski et al., 1997) .  ... 
doi:10.24230/kjiop.v34i1.1-25 fatcat:7atqmelsmfalvjvdatwwiicu3m

Video-Story Composition via Plot Analysis

Jinsoo Choi, Tae-Hyun Oh, In So Kweon
2016 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
Video-Story Composition via Plot Analysis Jinsoo Choi Tae-Hyun Oh In So Kweon  ... 
doi:10.1109/cvpr.2016.340 dblp:conf/cvpr/ChoiOK16 fatcat:xirguabhqbf35fbzb57regvdcq

Corporate Social Responsibility Reporting in the Casino Industry: A Content Analysis

Heather Markham Kim, Kyuha Cho, Younggin Choi, Junghoon (Jay) Lee, Jinsoo Hwang
2021 Sustainability  
The purpose of this study is to provide preliminary exploration of how corporate social responsibility issues are currently reported and communicated by US major casino companies. Empirical evidence is drawn from standalone CSR reports and websites of the 30 largest casino companies in the US, which is explored through content analysis methodology. This paper finds that there are substantial variations in the reporting and communication of CSR issues by casino companies. While casino companies
more » ... ainly rely on websites to communicate CSR issues, 23% of the sample companies use standalone CSR reports. Whereas websites tend to provide relatively limited information about CSR activities, standalone CSR reports have a tendency to cover comprehensive CSR themes in a detailed manner. In addition, the content analysis based on website reveals that while most casino companies tend to be passive about responsible gaming, some are trying to behave more actively and responsibly about this issue.
doi:10.3390/su132011185 fatcat:4fobhqmmyrgibnziafnq6trvre

Contextually Customized Video Summaries via Natural Language [article]

Jinsoo Choi, Tae-Hyun Oh, In So Kweon
2018 arXiv   pre-print
The best summary of a long video differs among different people due to its highly subjective nature. Even for the same person, the best summary may change with time or mood. In this paper, we introduce the task of generating customized video summaries through simple text. First, we train a deep architecture to effectively learn semantic embeddings of video frames by leveraging the abundance of image-caption data via a progressive and residual manner. Given a user-specific text description, our
more » ... lgorithm is able to select semantically relevant video segments and produce a temporally aligned video summary. In order to evaluate our textually customized video summaries, we conduct experimental comparison with baseline methods that utilize ground-truth information. Despite the challenging baselines, our method still manages to show comparable or even exceeding performance. We also show that our method is able to generate semantically diverse video summaries by only utilizing the learned visual embeddings.
arXiv:1702.01528v3 fatcat:fm7geb3efzc7fiile3w7ebslwe
« Previous Showing results 1 — 15 out of 243 results