Seoul, South Korea
Seoul, South Korea

Time filter

Source Type

Kim S.,Korea Advanced Institute of Science and Technology | Tai Y.-W.,Korea Advanced Institute of Science and Technology | Kim S.J.,SUNY Korea | Brown M.S.,National University of Singapore | Matsushita Y.,Microsoft
Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition | Year: 2012

This paper investigates the role that nonlinear camera response functions (CRFs) have on image deblurring. In particular, we show how nonlinear CRFs can cause a spatially invariant blur to behave as a spatially varying blur. This can result in noticeable ringing artifacts when deconvolution is applied even with a known point spread function (PSF). In addition, we show how CRFs can adversely affect PSF estimation algorithms in the case of blind deconvolution. To help counter these effects, we introduce two methods to estimate the CRF directly from one or more blurred images when the PSF is known or unknown. While not as accurate as conventional CRF estimation algorithms based on multiple exposures or calibration patterns, our approach is still quite effective in improving deblurring results in situations where the CRF is unknown. © 2012 IEEE.


Kim S.J.,SUNY Korea | Lin H.T.,National University of Singapore | Lu Z.,National University of Singapore | Susstrunk S.,Ecole Polytechnique Federale de Lausanne | And 2 more authors.
IEEE Transactions on Pattern Analysis and Machine Intelligence | Year: 2012

We present a study of in-camera image processing through an extensive analysis of more than 10,000 images from over 30 cameras. The goal of this work is to investigate if image values can be transformed to physically meaningful values, and if so, when and how this can be done. From our analysis, we found a major limitation of the imaging model employed in conventional radiometric calibration methods and propose a new in-camera imaging model that fits well with today's cameras. With the new model, we present associated calibration procedures that allow us to convert sRGB images back to their original CCD RAW responses in a manner that is significantly more accurate than any existing methods. Additionally, we show how this new imaging model can be used to build an image correction application that converts an sRGB input image captured with the wrong camera settings to an sRGB output image that would have been recorded under the correct settings of a specific camera. © 2012 IEEE.


Oh D.-S.,Chungnam National University | Phillips F.,Yuan Ze University | Phillips F.,State University of New York at Stony Brook | Park S.,SUNY Korea | Lee E.,Technopolis Development Center
Technovation | Year: 2016

Publications pushing the "innovation ecosystem" meme have added valuable dimensions to the economic development discussion. The phrase has captured the imagination of policy makers and has motivated public initiatives of substantial magnitude. This paper reviews the concept of innovation ecosystems as it is set forth in the academic and trade literature, and asks, "What is gained from adding 'eco-' to our treatment of national and regional innovation systems?". The answer is, "Very little, and the risks outweigh the benefits." Innovation ecosystem is not yet a clearly defined concept, much less a theory. Moreover, the idea carries pitfalls, notably its over-emphasis on market forces, and its flawed analogy to natural ecosystems.The prospect that the phrase "innovation ecosystem" is here to stay, in investment and economic development circles, implies a research gap, and indicates caution in using the phrase in rigorous research. The paper describes the gap, indicates directions for bridging it, and offers recommendations for prudent use of "ecosystem" terminology. © 2016 Elsevier Ltd.


Kim J.,SUNY Korea | Choe Y.,Texas A&M University | Mueller K.,SUNY Korea
Procedia Computer Science | Year: 2015

Integrating semantic features into parse trees is an active research topic in open-domain natural language processing (NLP). We study six different parse tree structures enriched with various semantic features for determining entity relations in clinical notes using a tree kernel-based relation extraction system. We used the relation extraction task definition and the dataset from the popular 2010 i2b2/VA challenge for our evaluation. We found that the parse tree structure enriched with entity type suffixes resulted in the highest F1 score of 0.7725 and was the fastest. In terms of reducing the number of feature vectors in trained models, the entity type feature was most effective among the semantic features while adding semantic feature node was better than adding feature suffixes to the labels. Our study demonstrates that parse tree enhancements with semantic features are effective for clinical relation extraction. © The Authors. Published by Elsevier B.V.


Betz F.,SUNY Korea | Betz F.,Portland State University
Journal of Sustainable Development | Year: 2015

Two important problems in models for a sustainable economy are (1) how to connect environmental models with societal models and (2) how to model the impact of societal changes upon an environment. For the first challenge, the Leontief economic model has been transformed from a vector to a tensor equation, as a model for integrating regional descriptions of an economy and its local environment. However, this is a static model of a steady-state (stasis) economy during a given period; and here we address the second challenge of modeling the impact of societal change upon an economy and its environment. We apply the modeling of societal change in the historical case of Australia's Mining Rent Tax of 2012-2014. © 2015 by the authors.


Pi J.,SUNY Korea | De P.,SUNY Korea | Mueller K.,SUNY Korea
Proceedings of the International Conference on Parallel and Distributed Systems - ICPADS | Year: 2013

We investigate the strength of patterns as secret signatures in Android's pattern based authentication mechanism. Parallelism of GPU is exploited to exhaustively search for the secret pattern. Typically, searching for a pattern, composed of a number of nodes and edges, requires an exhaustive search for the pattern. In this work, we show that the use of GPU can speed up the graph search, hence the pattern password, through parallelization. Preliminary results on cracking the Android pattern based passwords shows that the technique can be used as the basis to implement a tool that can check the strength of a pattern based password and thereby recommend strong patterns to the user. © 2013 IEEE.


Galdi G.P.,University of Pittsburgh | Mazzone G.,University of Pittsburgh | Mohebbi M.,SUNY Korea
Advances in Mathematical Fluid Mechanics - Dedicated to Giovanni Paolo Galdi on the Occasion of His 60th Birthday | Year: 2016

In this paper we investigate the existence of time-periodic motions of a system constituted by a rigid body with an interior cavity completely filled with a viscous liquid, and subject to a time-periodic external torque acting on the rigid body. We then show that the system of equations governing the motion of the coupled system liquid-filled rigid body, has at least one corresponding time-periodic weak solution. Furthermore if the size of the torque is below a certain constant, the weak solution is in fact strong. © Springer Basel 2016.


Im T.,SUNY Korea | De P.,SUNY Korea
Lecture Notes in Electrical Engineering | Year: 2016

Sensors on smartphones are used to enable different locationing techniques. Matching user generated images against a geo-tagged image database is a simple yet intuitive locationing technique. However, it is resource intensive and may take long to match. In this work, we present a lightweight positioning technique that uses outdoor images for locationing, but simplifies the matching process by using texts appearing in an image for locationing. The key intuition behind this work is that a collection of nearby textual signs appearing in a captured image forms a bag-of-words that provides a unique signature for a location. We design and implement a cloud-based positioning application, called WhereAmI, that overcomes the limitations of OCR on smartphones using a human-in-the-loop approach, as well as, offloading parts of the process to cloud server to save energy consumption. The experimental results comparing other locationing techniques show that WhereAmI can be an alternative positioning technique to GPS in terms of accuracy, energy efficiency, latency, and data usage. © Springer Science+Business Media Singapore 2016.


Tak N.H.,KRISS | Bahng G.W.,SUNY Korea | Kim T.G.,KIMMS
XXI IMEKO World Congress "Measurement in Research and Industry" | Year: 2015

The indicating device of a Leeb hardness tester isan important device that determines the performance of tester. As such, the indicating device needs verification just like the depth-measuring system of a Rockwell hardness tester. This paper presents the verification results of indicating devices from different manufacturers. For the verification, a calibrator to output the reference voltage signal was designed and used for the test. The results showed that the maximum deviation was 24 HLD.


Bhattcharya A.,SUNY Korea | De P.,SUNY Korea
Proceedings - Adaptive Resource Management and Scheduling for Cloud Computing: 3rd Internatioal Workshop, ARMS-CC 2016, Held in Conjunction with ACM Symposium on Principles of Distributed Computing, PODC 2016 | Year: 2016

Mobile devices like smartphones can augment their low-power processors by offloading portions of mobile applications to cloud servers. However, offloading to cloud data centers has a high network latency. To mitigate the problem of network latency, recently offloading to computing resources lying within the user's premises, such as network routers, tablets or laptop has been proposed. In this paper, we determine the devices whose processors have sufficient power to act as servers for computation offloading. We perform trace-driven simulation of SPECjvm2008 benchmarks to study the performance using different hardware. Our simulation shows that offloading to current state-of-the-art processors of user devices can improve performance of mobile applications. We find that offloading to user's own laptop reduces finish time of benchmark applications by 10%, compared to offloading to a commercial cloud server. © 2016 Copyright held by the owner/author(s).

Loading SUNY Korea collaborators
Loading SUNY Korea collaborators