TOHOKU UNIVERSITY Researchers

Details of the Researcher

Home

日本語 English

Jinhee Chun

Section

Graduate School of Information Sciences

Job title

Associate Professor

Degree

Ph.D. (Tohoku University)

researchmap

https://researchmap.jp/jinhee_chun

J-GLOBAL ID

200901048058544141

Education 2

Tohoku University　Graduate School, Division of Information Science　Department of System Information Sciences

- 2006/03/31
Tohoku University　Graduate School, Division of Information Science　Department of System Information Sciences

- 2003/03/31

Committee Memberships 1

電子情報通信学会コンピュテーション研究会　幹事

2012/06 - 2014/05

Professional Memberships 4

日本データベース学会
電子情報通信学会
情報処理学会
ACM

Research Interests 3

Computational Theory
Algorithm
Data Mining

Research Areas 2

Informatics / Intelligent informatics / 知識発見とデータマイニング
Informatics / Information theory / 計算幾何学

Awards 2

石田實記念財団研究奨励賞

2015/11/27　石田實記念財団　計算幾何を用いたイメージ検索モデルの構築と知識抽出
情報科学技術フォーラム(FIT)2002ヤングリサーチャー賞

2003/09/11　情報処理学会

Papers 52

Post-hoc Global Explanation using Hypersphere Sets. Peer-reviewed

Kohei Asano, Jinhee Chun

Proceedings of the 14th International Conference on Agents and Artificial Intelligence　236-243　2022
Publisher: SCITEPRESS
DOI： 10.5220/0010819100003116 　
Post-hoc Explanation using a Mimic Rule for Numerical Data. Peer-reviewed

Kohei Asano, Jinhee Chun

Proceedings of the 13th International Conference on Agents and Artificial Intelligence　768-774　2021
Publisher: SCITEPRESS
DOI： 10.5220/0010238907680774 　
CapsuleNet for Micro-Expression Recognition Peer-reviewed

Nguyen Van Quang, Jinhee Chun, Takeshi Tokuyama

2019 14th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2019)　1-7　2019/05
Publisher: IEEE
DOI： 10.1109/fg.2019.8756544 　
Consistent Digital Curved Rays and Pseudoline Arrangements. Peer-reviewed

Jinhee Chun, Kenya Kikuchi, Takeshi Tokuyama

27th Annual European Symposium on Algorithms(ESA)　32-16　2019
Publisher: Schloss Dagstuhl - Leibniz-Zentrum für Informatik
DOI： 10.4230/LIPIcs.ESA.2019.32 　
Model-Agnostic Explanations for Decisions Using Minimal Patterns Peer-reviewed

Kohei Asano, Jinhee Chun, Atsushi Koike, Takeshi Tokuyama

Artificial Neural Networks and Machine Learning – ICANN 2019: Theoretical Neural Computation　241-252　2019
Publisher: Springer International Publishing
DOI： 10.1007/978-3-030-30487-4_19 　

ISSN： 0302-9743

eISSN： 1611-3349
Large-Scale Taxonomy Problem: a Mixed Machine Learning Approach. Peer-reviewed

Quentin Labernia, Yashio Kabashima, Michimasa Irie, Toshiyuki Oike, Kohei Asano, Jinhee Chun, Takeshi Tokuyama

The SIGIR 2018 Workshop On eCommerce co-located with the 41st International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR 2018)(eCOM@SIGIR)　2018
Publisher: CEUR-WS.org
Gap-planar graphs. Peer-reviewed

Sang Won Bae, Jean-François Baffier, Jinhee Chun, Peter Eades, Kord Eickmeyer, Luca Grilli 0001, Seok-Hee Hong, Matias Korman, Fabrizio Montecchiani, Ignaz Rutter, Csaba D. Tóth

Theoretical Computer Science　745　36-52　2018

DOI： 10.1016/j.tcs.2018.05.029 　
Gap-planar Graphs. Peer-reviewed

Sang Won Bae, Jean-François Baffier, Jinhee Chun, Peter Eades, Kord Eickmeyer, Luca Grilli 0001, Seok-Hee Hong, Matias Korman, Fabrizio Montecchiani, Ignaz Rutter, Csaba D. Tóth

CoRR　abs/1708.07653　2017
Improving Thai Optical Character Recognition Using Circular-Scan Histogram. Peer-reviewed

Natsuda Kaothanthong, Thanaruk Theeramunkong, Jinhee Chun

14th IAPR International Conference on Document Analysis and Recognition(ICDAR)　567-572　2017
Publisher: IEEE
DOI： 10.1109/ICDAR.2017.98 　
Gap-Planar Graphs. Peer-reviewed

Sang Won Bae, Jean-François Baffier, Jinhee Chun, Peter Eades, Kord Eickmeyer, Luca Grilli 0001, Seok-Hee Hong, Matias Korman, Fabrizio Montecchiani, Ignaz Rutter, Csaba D. Tóth

Graph Drawing and Network Visualization - 25th International Symposium　531-545　2017
Publisher: Springer
DOI： 10.1007/978-3-319-73915-1_41 　
Folding Free-Space Diagrams: Computing the Fréchet Distance between 1-Dimensional Curves (Multimedia Contribution). Peer-reviewed

Kevin Buchin, Jinhee Chun, Maarten Löffler, Aleksandar Markovic 0001, Wouter Meulemans, Yoshio Okamoto, Taichi Shiitada

33rd International Symposium on Computational Geometry(SoCG)　64-5　2017
Publisher: Schloss Dagstuhl - Leibniz-Zentrum für Informatik
DOI： 10.4230/LIPIcs.SoCG.2017.64 　
Distance interior ratio: A new shape signature for 2D shape retrieval Peer-reviewed

Natsuda Kaothanthong, Jinhee Chun, Takeshi Tokuyama

PATTERN RECOGNITION LETTERS　78　14-21　2016/07

DOI： 10.1016/j.patrec.2016.03.029 　

ISSN： 0167-8655

eISSN： 1872-7344
A Reliable Top-k Dominating Query on Uncertain Big Database Peer-reviewed

Xiang Yu, Jinhee Chun, Takeshi Tokuyama

Proceedings of the 18th Korea-Japan Joint Workshop on Algorithms and Computation (WAAC2015)　59-64　2015
Reliable Top-k Dominating Query on Uncertain Big Data with x-Relation Peer-reviewed

Xiang Yu, Jinhee Chun, Takeshi Tokuyama

Proceedings of the Young Researcher Workshop on Automata, Languages and Programming (YR-ICALP2015)　33-41　2015
Base-object location problems for base-monotone regions Peer-reviewed

Jinhee Chun, Takashi Horiyama, Takehiro Ito, Natsuda Kaothanthong, Hirotaka Ono, Yota Otachi, Takeshi Tokuyama, Ryuhei Uehara, Takeaki Uno

THEORETICAL COMPUTER SCIENCE　555　71-84　2014/10

DOI： 10.1016/j.tcs.2013.11.030 　

ISSN： 0304-3975

eISSN： 1879-2294
A Unified View to Greedy Geometric Routing Algorithms in Ad Hoc Networks Peer-reviewed

Jinhee Chun, Akiyoshi Shioura, Truong Minh Tien, Takeshi Tokuyama

IEICE TRANSACTIONS ON FUNDAMENTALS OF ELECTRONICS COMMUNICATIONS AND COMPUTER SCIENCES　E97A　(6)　1220-1230　2014/06

DOI： 10.1587/transfun.E97.A.1220 　

ISSN： 1745-1337
Shape Description using Classified Distances Peer-reviewed

Jinhee Chun, Natsuda Kaothanthong, Takeshi Tokuyama

Proceedings of the 7th Annual Meeting of Asian Association for Algorithms and Computation (AAAC2014)　12-12　2014
A unified view to greedy geometric routing algorithms in ad hoc networks Peer-reviewed

Jinhee Chun, Akiyoshi Shioura, Truong Minh Tien, Takeshi Tokuyama

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)　7718　54-65　2013
Publisher: Springer
DOI： 10.1007/978-3-642-36092-3-7 　

ISSN： 0302-9743 1611-3349
Classified-distance based shape descriptor for application to image retrieval Peer-reviewed

Jinhee Chun, Natsuda Kaothanthong, Takeshi Tokuyama

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)　8048　(2)　1-8　2013
Publisher: Springer
DOI： 10.1007/978-3-642-40246-3_1 　

ISSN： 0302-9743 1611-3349
A Multiple Instance Learning Approach to Image Annotation with Saliency Map Peer-reviewed

Tran Phuong Nhung, Cam-Tu Nguyen, Jinhee Chun, Ha Vu Le, Takeshi Tokuyama

Proceedings of the International Conference on Knowledge Discovery and Information Retrieval and the International Conference on Knowledge Management and Information Sharing (KDIR/KMIS 2013)　150-159　2013
Publisher: SciTePress
DOI： 10.5220/0004543901520159 　
Space-Efficient and Data-Sensitive Polygon Reconstruction Algorithms from Visibility Angle Information Peer-reviewed

Jinhee Chun, Ricardo Garcia de Gonzalo, Takeshi Tokuyama

ALGORITHMS AND COMPUTATION　8283　22-32　2013

DOI： 10.1007/978-3-642-45030-3_3 　

ISSN： 0302-9743
Shape Description using Classified Distances Peer-reviewed

Jinhee Chun, Natsuda Kaothanthong, Takeshi Tokuyama

Proceedings of the 6th Annual Meeting of Asian Association for Algorithms and Computation (AAAC2013)　29-29　2013
Base location problems for base-monotone regions Peer-reviewed

Jinhee Chun, Takashi Horiyama, Takehiro Ito, Natsuda Kaothanthong, Hirotaka Ono, Yota Otachi, Takeshi Tokuyama, Ryuhei Uehara, Takeaki Uno

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)　7748　53-64　2013
Publisher: Springer
DOI： 10.1007/978-3-642-36065-7_7 　

ISSN： 0302-9743 1611-3349
Correspondence Finder using Classified Distance Distribution for Efficient Shape Retrieval Peer-reviewed

Jinhee Chun, Natsuda Kaothanthong, Takeshi Tokuyama

Proceedings of the 6th Korea-Japan Joint Workshop on Algorithms and Computation (WAAC 2013)　25-38　2013
Algorithms for computing the maximum weight region decomposable into elementary shapes Invited Peer-reviewed

Jinhee Chun, Natsuda Kaothanthong, Ryosei Kasai, Matias Korman, Martin Noellenburg, Takeshi Tokuyama

COMPUTER VISION AND IMAGE UNDERSTANDING　116　(7)　803-814　2012/07

DOI： 10.1016/j.cviu.2012.03.003 　

ISSN： 1077-3142

eISSN： 1090-235X
Optimal grid decompositions for maximum weight region computation with application to image segmentation Peer-reviewed

Jinhee Chun, Natsuda Kaothanthong, Hiromi Takahashi

Computational Geometry: Young Researchers Forum　2012
A Unified View to Greedy Routing Algorithms in Ad Hoc Networks Peer-reviewed

Jinhee Chun, Akiyoshi Shioura, Truong Minh Tien, Takeshi Tokuyama

Proceedings of the 15th Korea-Japan Joint Workshop on Algorithms and Computation(WAAC2012)　33-40　2012
How to Cut a Complicated Figure by Using Scissors? Computing the Maximum Weight Region Consisting of Base monotone Regions Peer-reviewed

Jinhee Chun, Natsuda Kaothanthong, Hiromi Takahashi, Takeshi Tokuyama

Proceedings of the 5th Annual Meeting of Asian Association for Algorithms and Computation(AAAC2012)　2012
Image Retrieval System Using Distance-Based Shape Recognition Peer-reviewed

Jinhee Chun, Natsuda Kaothanthong, Yoshiki Ota, Takeshi Tokuyama

Proceedings of the 14th Korea-Japan Joint Workshop on Algorithms and Computation (WAAC2011)　2011/07
Distance-Based Shape Invariants for Image Retrieval Peer-reviewed

Jinhee Chun, Natsuda Kaothanthong, Yoshiki Ota, Takeshi Tokuyama

The 4th Annual Meeting of the Asian Association for Algorithms and Computation(AAAC2011)　2011/04
Effect of Corner Information in Simultaneous Placement of K Rectangles and Tableaux Invited Peer-reviewed

Shinya Anzai, Jinhee Chun, Ryosei Kasai, Matias Korman, Takeshi Tokuyama

Discrete Mathematics, Algorithms and Applications　2　(4)　527-537　2010/12

DOI： 10.1142/S1793830910000863 　
Algorithms for Image Segmentation Based on Geometric Approach Peer-reviewed

Shinya Anzai, Jinhee Chun, Ryosei Kasai, Matias Korman, Takeshi Tokuyama

Proceedings of Asian Association for Algorithms and Computation 2010　2010/04
Distance Trisector of a Segment and a Point Peer-reviewed

Jinhee Chun, Yuji Okada, Takeshi Tokuyama

Interdisciplinary Information Sciences　16　(1)　119-126　2010/03
Publisher: Tohoku University
DOI： 10.4036/iis.2010.119 　

ISSN： 1340-9050
Effect of corner information in simultaneous placement of K rectangles and tableaux Peer-reviewed

Shinya Anzai, Jinhee Chun, Ryosei Kasai, Matias Korman, Takeshi Tokuyama

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)　6196　235-243　2010
Publisher: Springer
DOI： 10.1007/978-3-642-14031-0_27 　

ISSN： 0302-9743 1611-3349
Consistent Digital Rays Invited Peer-reviewed

Jinhee Chun, Matias Korman, Martin Noellenburg, Takeshi Tokuyama

DISCRETE & COMPUTATIONAL GEOMETRY　42　(3)　359-378　2009/10

DOI： 10.1007/s00454-009-9166-2 　

ISSN： 0179-5376
Algorithms for Optimal Segmentation of Regions Decomposable into Basic Shapes Peer-reviewed

Jinhee Chun, Ryosei Kasai, Matias Korman, Takeshi Tokuyama

Proceedings of the 11th Japan-Korea Joint Workshop on Algorithms and Computation　138-145　2009/07
Algorithms for Computing the Maximum Weight Region Decomposable into Elementary Shapes Peer-reviewed

Jinhee Chun, Ryosei Kasai, Matias Korman, Takeshi Tokuyama

ALGORITHMS AND COMPUTATION, PROCEEDINGS　5878　1166-1174　2009

DOI： 10.1007/978-3-642-10631-6_117 　

ISSN： 0302-9743
Finding the maximum union of closures is NP-hard, even for trees Peer-reviewed

Jinhee Chun, Matias Korman, Martin Noellenburg, Takeshi Tokuyama

Proceedings of the 11th Japan-Korea Joint Workshop on Algorithms and Computation　139-144　2008/07
Digital Star Shapes and Their Applications Peer-reviewed

Jinhee Chun, Matias Korman, Martin Nöllenburg, Takeshi Tokuyama

Proceedings of the Asian Association for Algorithms and Computation　2008/04
Consistent Digital Rays Peer-reviewed

Jinhee Chun, Matias Korman, Martin Noellenburg, Takeshi Tokuyama

PROCEEDINGS OF THE TWENTY-FOURTH ANNUAL SYMPOSIUM ON COMPUTATIONAL GEOMETRY (SGG'08)　355-364　2008
Consistent Digital Rays Peer-reviewed

Jinhee Chun, Matias Korman, Martin Noellenburg, Takeshi Tokuyama

PROCEEDINGS OF THE TWENTY-FOURTH ANNUAL SYMPOSIUM ON COMPUTATIONAL GEOMETRY (SGG'08)　355-364　2008

DOI： 10.1145/1377676.1377737 　
Distance trisector of segments and zone diagram of segments in a plane Peer-reviewed

Jinhee Chun, Yuji Okada, Takeshi Tokuyama

ISVD 2007: THE 4TH INTERNATIONAL SYMPOSIUM ON VORONOI DIAGRAMS IN SCIENCE AND ENGINEERING 2007, PROCEEDINGS　66-+　2007

DOI： 10.1109/ISVD.2007.19 　
Linear time algorithm for approximating a curve by a single-peaked curve Invited Peer-reviewed

J Chun, K Sadakane, T Tokuyama

ALGORITHMICA　44　(2)　103-115　2006/02

DOI： 10.1007/s00453-005-1201-y 　

ISSN： 0178-4617
Efficient algorithms for constructing a pyramid from a terrain Peer-reviewed

J Chun, K Sadakane, T Tokuyama

IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS　E89D　(2)　783-788　2006/02

DOI： 10.1093/ietisy/e89-d.2.783 　

ISSN： 0916-8532
Peak-Reducing Fitting of a Curve under the $L_p$ Metric Peer-reviewed

Jinhee Chun, Sadakane, Takeshi Tokuyama, Masato yuki

Interdisciplinary Information Sciences　11　(2)　191-197　2005/09/16
Publisher: Tohoku University
DOI： 10.4036/iis.2005.191 　

ISSN： 1340-9050

More details Close

Given a function y=f(x) in one variable, we consider the problem of computing a k-peaked curve y=φ(x) minimizing the L_p distance between them. In other words, φ(x) has at most k local peaks and minimizes the area bounded by the curves f(x) and φ(x). This gives extension of the authors' previous work [5] on the unimodal (i.e., single-peaked) approximation for the L₂ distance.
Construction of Expert Guided Decision Tree on Numeric Database Peer-reviewed

Jinhee Chun, Sadakane, Takeshi Tokuyama

Proc. WAAC2005　27-34　2005/08
Efficient algorithms for approximating a multi-dimensional voxel terrain by a unimodal terrain Peer-reviewed

DZ Chen, JH Chun, N Katoh, T Tokuyama

COMPUTING AND COMBINATORICS, PROCEEDINGS　3106　238-248　2004

DOI： 10.1007/b99243 　

ISSN： 0302-9743
Linear Time Algorithm for Approximating a Curve by a Single-Peaked Curve Peer-reviewed

Jinhee Chun, Sadakane, Takeshi Tokuyama

Proc. WAAC2003　116-123　2003/07
高次元ピラミッドを用いた数値属性結合ルールの生成とデータマイニングへの応用 Peer-reviewed

全眞嬉, Danny Z. Chen, 加藤直樹, 徳山豪

日本データベース学会Letters　2　(1)　83-86　2003/05
Linear time algorithm for approximating a curve by a single-peaked curve Peer-reviewed

JH Chun, K Sadakane, T Tokuyama

ALGORITHMS AND COMPUTATION, PROCEEDINGS　2906　6-15　2003

DOI： 10.1007/978-3-540-24587-2_3 　

ISSN： 0302-9743
地形図からの最適ピラミッドの構成アルゴリズム Peer-reviewed

全眞嬉, 加藤直樹, 徳山豪

情報技術レターズ　1　(LA-2)　3-4　2002/09
Efficient algorithms for constructing a pyramid from a terrain Peer-reviewed

JH Chun, K Sadakane, T Tokuyama

DISCRETE AND COMPUTATIONAL GEOMETRY　2866　108-117　2002

DOI： 10.1007/978-3-540-44400-8_11 　

ISSN： 0302-9743

Show all ︎Show first 5

Misc. 30

Model-Agnostic Explanations with Mimic Rules

ASANO Kohei, CHUN Jinhee

Proceedings of the Annual Conference of JSAI　JSAI2020　3E1GS203-3E1GS203　2020
Publisher: The Japanese Society for Artificial Intelligence
DOI： 10.11517/pjsai.jsai2020.0_3e1gs203 　

More details Close

Recently, a large number of machine learning models have been proposed. Unfortunately, such models often make black-box decisions that are not easy to explain the logical reasons to derive them. Therefore, it is important to develop a tool that automatically gives the reasons for the model’s decision. Some research tackle to solve this problem by approximating an explained model with an interpretable model such as a decision tree. Although these methods provide logical reasons for a model's decision, it sometimes occurs a wrong explanation. We propose a novel model-agnostic explanation method with the rule models that we call mimic rules. Mimic rules are an interpretable model and have the same outputs to an explained model. We give a comparison of our method to previous methods, and we show that our method often improves local fidelity.
整合的なデジタル曲線族の構築法

菊地, 賢也, 全, 眞嬉, 徳山, 豪

第81回全国大会講演論文集　2019　(1)　181-182　2019/02/28

More details Close

デジタル空間ではパターンが点の集合によって表され，このように幾何学的物体をデジタル空間に表すことはコンピュータグラフィックスや計算幾何学などの分野において基本的な問題である．ここで，デジタル化されたパターン同士が交差する状態はユークリッド幾何学での状態と同様になることが望ましい．先行研究より，原点を端点とする半直線の族はデジタル空間上で葉を持たないグリッドの全域木を形成し，デジタル化された各半直線と元のユークリッド空間上の半直線の距離はΘ(log<l>n<l>)であることが示された．本研究はこの理論を一般化し，原点を通る曲線族へと拡張する．
制約付き最適化を用いた非負PNU半教師学習による過学習の抑制

石川, 周, 全, 眞嬉, 徳山, 豪

第81回全国大会講演論文集　2019　(1)　47-48　2019/02/28

More details Close

近年の通信技術や情報サービスの発展と多様化により，膨大な量と多種にわたるデータが利用可能となった．それらを用いることで，様々な分野や目的に対して機械学習によるパターン認識の活用が期待されている．本研究では，機械学習における分類問題を取り扱う．ディープラーニングと呼ばれるような多層パーセプトロンや畳み込みニューラルネットワークなどの複雑なモデルに対して過学習を抑えられ，学習データの生成にかかるコスト削減の実現を目指す半教師あり学習の方法について取り組む．具体的なアプローチとして，先行研究であるPNU LearningにNon-Negative PU Learningを組み込むことでこれを実現する．
Model-agnostic Explainer using Feature Patterns

ASANO Kohei, CHUN Jinhee, TOKUYAMA Takeshi

Proceedings of the Annual Conference of JSAI　JSAI2019　2Q3J204-2Q3J204　2019
Publisher: The Japanese Society for Artificial Intelligence
DOI： 10.11517/pjsai.jsai2019.0_2q3j204 　

More details Close

Recently, high performance, though very complex, machine learning models have been proposed. Being able to interpret such black boxed decision making models is clearly critical for sensible tasks. In this paper, we propose a model-agnostic explanation method that details the prediction behavior of any models using feature patterns. Since the relationship between features is still unclear in the previous works, we focus on a combination of these features. We propose an algorithm which finds several minimal feature patterns that lead target prediction sufficiently using hill climbing search. In our experiments, we aim at measuring the faithfulness of our explanation, thus apply our method to sentiment analysis dataset and evaluate the faithfulness using two metrics: recall and precision. At last, we demonstrate the benefit of our method based on some image classification use cases with a black-box model.
D-031 Efficient Top-k Dominating Query on Uncertain Database

Yu Xiang, Tokuyama Takeshi, Chun Jinhee

14　(2)　143-144　2015/08/24
Publisher: Forum on Information Technology

More details Close

We study the problem of ranking queries on uncertain databases where objects are mutually exclusive. We utilize the top-k dominating query with x-Relation data model, and propose novel dominance criteria to return reliable top-k answers. Moreover, we present pruning rules to reduce the computation. The experiments show that our ranking method is more reliable than the uncertain top-k query method. The runtime and space are also theoretically promising.
Space-efficient Polygon Reconstruction from Angle Information

CHUN Jinhee, GARCIA DE GONZARO Ricardo, TOKUYAMA Takeshi

IEICE technical report. Theoretical foundations of Computing　113　(14)　27-31　2013/04/24
Publisher: The Institute of Electronics, Information and Communication Engineers
ISSN： 0913-5685

More details Close

We propose a linear working space algorithm for reconstructing a simple polygon from the visible angle information at vertices.
A Unified View to Greedy Geometric Routing Algorithms in Ad Hoc Networks

GHUN Jinhee, SHIOURA Akiyoshi, MINH TIEN Truong, TOKUYAMA Takeshi

IEICE technical report. Theoretical foundations of Computing　112　(272)　1-8　2012/10/24
Publisher: The Institute of Electronics, Information and Communication Engineers
ISSN： 0913-5685

More details Close

The main aim of this paper is to give a unified view to greedy geometric routing algorithms in ad hoc networks. For this, we firstly present a general form of greedy routing algorithm using a class of objective functions which are invariant under congruent transformations of a point set. We show that some known greedy routing algorithms such as Greedy Routing, Compass Routing, and Midpoint Routing can be regarded as special cases of the generalized greedy routing algorithm. In addition, inspired by the unified view of greedy routing, we propose three new greedy routing algorithms. We then derive a sufficient condition for our generalized greedy routing algorithm to guarantee packet delivery on every Delaunay graph. This condition makes it easier to check whether a given routing algorithm guarantees packet delivery, and it is closed under convex linear combination of objective functions. We show that Greedy Routing, Midpoint Routing, and the three new greedy routing algorithms proposed in this paper satisfy the sufficient condition, i.e., they guarantee packet delivery on Delaunay graphs, and then discuss comparison of these methods.
DS-1-11 An Algorithm for Optimally Locating Baselines using Quad Decomposition

Chun Jinhee, Kaothanthong Natsuda, Otachi Yota, Takahashi Hiromi, Tokuyama Takeshi

Proceedings of the IEICE General Conference　2012　(1)　"S-21"-"S-22"　2012/03/06
Publisher: The Institute of Electronics, Information and Communication Engineers

More details Close

In this research, we want to improve the segmented result obtained from the room-edge algorithm. We present an algorithm for placing baselines which gives the optimal room-edge region. The baselines are located using a quad decomposition which recursively decompose a pixel grid P into 4^l rectangular pixel grid for given l levels. The optimal room-edge region of an m×n rectangular pixel grid G can be computed in O(mn^2) time. The optimal location for horizontal and vertical baselines of the given n×n pixel grid P can be computed in O(n^7) time using O(n^4) space. The experimental result is presented in this work.
Image Segmentation Using Maximum Weight Region

CHUN Jinhee, KAOTHANTHONG Natsuda, TOKUYAMA Takeshi

IEICE technical report. Theoretical foundations of Computing　111　(360)　61-67　2011/12/16
Publisher: The Institute of Electronics, Information and Communication Engineers
ISSN： 0913-5685

More details Close

In this paper, we present the image segmentation result using the algorithm for computing the Maximum Weight Region. An input monochrome image is represented by a weighted n × n pixel grid where each pixel corresponds to the intensity value of the input. The pixel grid is segmented into a set of disjoint elementary shapes using the algorithm proposed by Chun et al.. The quality of the segmented image can be improved by dividing the grid into (k+1)^2 cells consisting of N_t pixels using k baselines. Each cell can be computed using O(N_t^<1.5>) time and requires O(N_t) space.
タブローの最適配置問題 (計算機科学とアルゴリズムの数理的基礎とその応用)

安齋進也, 全眞嬉, 葛西亮生, コルマンマティアス, 徳山豪

数理解析研究所講究録　1744　93-98　2011/06
Publisher: 京都大学
ISSN： 1880-2818
Image Recognition and Retrieval by Using Distance Information (Mathematical Foundations and Applications of Computer Science and Algorithms)

Chun Jinhee, Kaothanthong Natsuda, Tokuyama Takeshi

RIMS Kokyuroku　1744　185-188　2011/06
Publisher: Kyoto University
ISSN： 1880-2818
Optimal Tableaux Placement Problem

2011　(2)　1-7　2011/02/28
Publisher: 情報処理学会
ISSN： 2186-2583
画像切り出しに対するアルゴリズムの提案 (アルゴリズムと計算機科学の数理的基盤とその応用)

安齋進也, 全眞嬉, 葛西亮生, コルマンマティアス, 徳山豪

数理解析研究所講究録　1691　195-199　2010/06
Publisher: 京都大学
ISSN： 1880-2818
Effect of Corner Information in Simultaneous Placement of k Rectangles and Tableaux

ANZAI Shinya, CHUN Jinhee, KASAI Ryosei, KORMAN Matias, TOKUYAMA Takeshi

IEICE technical report　110　(12)　33-38　2010/04/15
Publisher: The Institute of Electronics, Information and Communication Engineers
ISSN： 0913-5685

More details Close

We consider the optimization problem of finding k nonintersecting rectangles and tableaux in n × n pixel plane where each pixel has a real valued weight. We discuss existence of efficient algorithms if a corner point of each rectangle/tableau is specified.
On simultaneous optimal segmentation of digital objects

CHUN Jinhee, KASAI Ryosei, KORMAN Matias, TOKUYAMA Takeshi

IEICE technical report　109　(235)　1-8　2009/10/09
Publisher: The Institute of Electronics, Information and Communication Engineers
ISSN： 0913-5685

More details Close

Motivated from the image segmentation problem, we consider the problem of finding the maximum weight region with a shape decomposable into elementary shapes in n×n pixel grid where each pixel has a real valued weight. We give efficient algorithms for several interesting cases, extending our previous paper [13]. This shows a strong contrast with the NP-hardness results on finding the maximum weight union for the corresponding cases.
D-026 Automatic Taxonomy Generating Methods for Just-in-Time Web Advertising

Sakuraba Atsuyuki, Narita Ryota, Chun Jinhee, Tokuyama Takeshi

8　(2)　185-186　2009/08/20
Publisher: Forum on Information Technology
RA-006 Algorithms for computing the optimal image segmentation of nonintersecting union of base monotone regions

Chun Jinhee, Kasai Ryosei, Korman Matias, Tokuyama Takeshi

8　(1)　37-42　2009/08/20
Publisher: Forum on Information Technology
Algorithms for optimal segmentation of regions decomposable into basic shapes: 基本図形分割可能領域の最適切り出しアルゴリズム: Theoretical foundations of computing

カサイリョウセイ

109　(108)　23-30　2009/06/29
Publisher: The Institute of Electronics, Information and Communication Engineers
ISSN： 0913-5685
D-022 Automatic Web Search System Using Comments for Video Contents

Narita Ryouta, Chun Jinhee, Tokuyama Takeshi

7　(2)　93-94　2008/08/20
Publisher: Forum on Information Technology
Consistent digital rays: 数学的整合性を持つデジタル直線集合

107　(537)　39-46　2008/03/10
Publisher: The Institute of Electronics, Information and Communication Engineers
ISSN： 0913-5685
D_036 Automatic extraction and conversion of data from the WWW

Onuma Hiroaki, Chun Jinhee, Tokuyama Takeshi

5　(2)　83-84　2006/08/21
Publisher: Forum on Information Technology
Algorithms and Implementation of Peak-Reducing Fitting of a Curve

YUKI Masato, CHUN Jinhee, SADAKANE Kunihiko, TOKUYAMA Takeshi

IEICE technical report. Theoretical foundations of Computing　104　(642)　83-89　2005/01/21
Publisher: The Institute of Electronics, Information and Communication Engineers
ISSN： 0913-5685

More details Close

The problem of approximating complexity curve is basic operation for processing of geometry data. In this paper, We emphasize characteristic of output curve. We consider k-peaked curve φ that the distance between f and φ is minimized. We design algorithms for computing φ for L_∞-distance and L_1-distance.
データ分布の近似を用いたエキスパート付き決定木構築

全眞嬉

応用数理学会, 東北大, Sept. 2005　2005
F-050 Construction of Expert Guided Decision Tree on Numeric Database

Chun Jinhee, Kaneko Masanori, Sadakane Kunihiko, Tokuyama Takeshi

3　(2)　331-332　2004/08/20
Publisher: Forum on Information Technology
数値属性を用いたエキスパート付き決定木の構築

全眞嬉, 徳山豪

日本ソフトウェア学会研究会資料:第4回データマイニングワークショップ 29　21-24　2004
Higher - Dimensional Pyramid Construction Problem and Application to Data Mining

CHEN Danny Z, CHUN Jinhee, KATOH Naoki, TOKUYAMA Takeshi

IPSJ SIG Notes　2003　(3)　71-78　2003/01/20
Publisher: Information Processing Society of Japan (IPSJ)
ISSN： 0919-6072

More details Close

Optimized region rules developed by Fukuda et al. [9, 10] are effective tools for data mining in databases with numeric data. However, there are two drawbacks in the previous methods: (1) each rule can contain at most two numeric conditional attributes, and (2) the decision is made based only on whether a given data is inside or outside a region R, but not on the exact position of the data. In this paper, we propose a new method for removing these drawbacks. Indeed, by applying graph algorithmd, we give optimized numeric association rules with more than two attributes, and give layered-structure numeric association rules. Our method is also applicable to removal of exceptional data and data clustering.
How to reform a terrain into a pyramid

CHUN Jinhee, KATOH Naoki, TOKUYAMA Takeshi

IPSJ SIG Notes　2002　(88)　51-58　2002/09/19
Publisher: Information Processing Society of Japan (IPSJ)
ISSN： 0919-6072

More details Close

Given nonnegative valued functions ρ and μ in d variables, we consider the optimal pyramid maximizing the total parametric gain of ρ against μ. The pyramid can be considered as the optimal unimodal approximation of ρ relative to μ, and can be applied to hierarchical data segmentation. We study properties of the optimal pyramid, and design efficient algorithms for several cases mainly for d = 1 and 2.
大規模多品種購買データに対するデータマイニング

全眞嬉, 徳山, 豪, 藤井, 章博

第63回全国大会講演論文集　2001　(1)　191-192　2001/09/26
大学における情報技術教育事例--宮城大学事業構想学部

藤井章博, 岩間問, 全眞嬉

情報処理学会研究報告 = IPSJ SIG technical reports　2000　(117)　55-60　2000/12/15
Publisher: 一般社団法人情報処理学会
ISSN： 0919-6072
資料情報技術教育事例

藤井章博, 岩間問, 全眞嬉

宮城大学事業構想学部紀要 = Bulletin of Miyagi University, School of Project Design / 宮城大学事業構想学部研究紀要広報委員会編　(3)　139-148　2000
Publisher: 宮城大学事業構想学部
ISSN： 1344-607X

Show all ︎Show first 5

Presentations 45

A Reliable Top-k Dominating Query on Uncertain Big Database International-presentation

Xiang Yu, Jinhee Chun, Takeshi Tokuyama

The 18th Korea-Japan Joint Workshop on Algorithms and Computation (WAAC2015)　2015/08
Reliable Top-k Dominating Query on Uncertain Big Data with x-Relation International-presentation

Xiang Yu, Jinhee Chun, Takeshi Tokuyama

Young Researcher Workshop on Automata, Languages and Programming (YR-ICALP2015)　2015/08
Shape Description using Classified Distances International-presentation

Jinhee Chun, Natsuda Kaothanthong, Takeshi Tokuyama

Proceedings of the 7th Annual Meeting of Asian Association for Algorithms and Computation (AAAC2014)　2014/04
Space-Efficient and Data-Sensitive Polygon Reconstruction Algorithms from Visibility Angle Information International-presentation

Jinhee Chun, Ricardo G.Gonzalo, Takeshi Tokuyama

The 24th Annual International Symposium on Algorithms and Computation (ISAAC2013)　2013/12
Base Location Problems for Base-Monotone Regions International-presentation

Jinhee Chun, Takashi Horiyama, Takehiro Ito, Natsuda Kaothanthong, Hirotaka Ono, Yota Otachi, Takeshi Tokuyama, Ryuhei Uehara, Takeaki Uno

The 7th International Workshop on Algorithms and Computation(WALCOM 2013)　2013/12
A Multiple Instance Learning Approach to Image Annotation with Saliency Map International-presentation

Tran Phuong Nhung, Cam-Tu Nguyen, Jinhee Chun, Ha Vu Le, Takeshi Tokuyama

The International Conference on Knowledge Discovery and Information Retrieval and the International Conference on Knowledge Management and Information Sharing (KDIR/KMIS 2013)　2013/09
Classified-Distance Based Shape Descriptor for Application to Image Retrieval International-presentation

Jinhee Chun, Natsuda Kaothanthong, Takeshi Tokuyama

Proceedings of the 15th International Conference on Computer Analysis of Images and Patterns (CAIP2013)　2013/08
Correspondence Finder using Classified Distance Distribution for Efficient Shape Retrieval International-presentation

Jinhee Chun, Natsuda Kaothanthong, Takeshi Tokuyama

Proceedings of the 6th Korea-Japan Joint Workshop on Algorithms and Computation (WAAC 2013)　2013/08
Shape Description using Classified Distances International-presentation

Jinhee Chun, Natsuda Kaothanthong, Takeshi Tokuyama

Proceedings of the 6th Annual Meeting of Asian Association for Algorithms and Computation (AAAC2013)　2013/04
A Unified View to Greedy Geometric Routing Algorithms in Ad Hoc Networks International-presentation

Jinhee Chun, Akiyoshi Shioura, Truong Minh Tien, Takeshi Tokuyama

Algosensors2012　2012/09
A Unified View to Greedy Routing Algorithms in Ad Hoc Networks International-presentation

Jinhee Chun, Akiyoshi Shioura, Truong Minh Tien, Takeshi Tokuyama

The 15th Korea-Japan Joint Workshop on Algorithms and Computation(WAAC2012)　2012/08
Optimal grid decompositions for maximum weight region computation with application to image segmentation International-presentation

Jinhee Chun, Natsuda Kaothanthong, Hiromi Takahashi

Computational Geometry: Young Researchers Forum　2012/06
How to Cut a Complicated Figure by Using Scissors? Computing the Maximum Weight Region Consisting of Base monotone Regions International-presentation

Jinhee Chun, Natsuda Kaothanthong, Hiromi Takahashi, Takeshi Tokuyama

The 5th Annual Meeting of Asian Association for Algorithms and Computation(AAAC2012)　2012/04
Image Retrieval System Using Distance-Based Shape Recognition International-presentation

Jinhee Chun, Natsuda Kaothanthong, Yoshiki Ota, Takeshi Tokuyama

The 14th Korea-Japan Joint Workshop on Algorithms and Computation (WAAC2011)　2011/07
Distance-Based Shape Invariants for Image Retrieval International-presentation

Jinhee Chun, Natsuda Kaothanthong, Yoshiki Ota, Takeshi Tokuyama

The 4th Annual Meeting of the Asian Association for Algorithms and Computation(AAAC2011)　2011/04
Effect of Corner Information in Simultaneous Placement of K Rectangles and Tableaux International-presentation

Shinya Anzai, Jinhee Chun, Ryosei Kasai, Matias Korman, Takeshi Tokuyama

The 16th International Computing and Combinatorics Conference（COCOON’10）　2010/07
Algorithms for Image Segmentation Based on Geometric Approach International-presentation

Shinya Anzai, Jinhee Chun, Ryosei Kasai, Matias Korman, Takeshi Tokuyama

Asian Association for Algorithms and Computation 2010　2010/04
Algorithms for Computing the Maximum Weight Region Decomposable into Elementary Shapes International-presentation

Jinhee Chun, Ryosei Kasai, Matias Korman, Takeshi Tokuyama

The 20th International Symposium on Algorithms and Computation　2009/12
Algorithms for Optimal Segmentation of Regions Decomposable into Basic Shapes International-presentation

Jinhee Chun, Ryosei Kasai, Matias Korman, Takeshi Tokuyama

The 11th Japan-Korea Joint Workshop on Algorithms and Computation　2009/07
Consistent Digital Rays

Jinhee Chun, Matias Korman, Martin Nöllenburg, Takeshi Tokuyama

The 2 nd International Symposium on Information Electronics Systems　2008/07
Digital Star Shapes and Their Applications International-presentation

Jinhee Chun, Matias Korman, Martin Nöllenburg, Takeshi Tokuyama

Asian Association for Algorithms and Computation 2008　2008/04
Consistent digital rays

Jinhee Chun, Matias Korman, Martin Nöllenburg, Takeshi Tokuyama

コンピュテーション研究会　2008/03
Digital Star Shapes and Their Applications

Jinhee Chun, Matias Korman, Martin Nollenburg, Takeshi Tokuyama

情報処理学会114回アルゴリズム研究会　2007/09
Distance Trisector of Segments and Zone Diagram of Segments in a Plane International-presentation

Jinhee Chun, Yuji Okada, Takeshi Tokuyama

4th International Symposium on Voronoi Diagrams in Science and Engineering(ISVD 2007)　2007/07
Webデータの自動抽出とデータ変換

小沼寛明, 全眞嬉, 徳山豪

FIT2006報科学技術フォーラム　2006/09
Construction of Expert Guided Decision Tree on Numeric Database International-presentation

Jinhee Chun, Kunihiko Sadakane, Takeshi Tokuyama

8th Japan-Korea Workshop on Algorithm and Computation (WAAC05)　2005/08
最適ピラミッドを用いた決定木構築

全眞嬉, 定兼邦彦, 徳山豪

LAシンポジウム　2005/07
データ分布の近似を用いたエキスパート付き決定木構築

全眞嬉, 定兼邦彦, 徳山豪

日本応用数理学会JSIAM05　2005/02
曲線ピーク削減アルゴリズムの考察と実装

結城匡人, 全眞嬉, 定兼邦彦, 徳山豪

電子情報通信学会コンピュテーション研究会　2005/01
Peak-Reducing Fitting of a Curve under the L1 Metric International-presentation

Jinhee Chun, Kunihiko Sadakane, Takeshi Tokuyama

Japan Conference on Discrete and Computational Geometry 2004（JCDCG04)　2004/10
Construction of Expert Guided Decision Tree on Numeric Database International-presentation

Jinhee Chun,Kunihiko, Sadakane, Takeshi Tokuyama

The 2nd Student-Organizing International Mini-Conference on Information Electronics System(SOIM-COE04)　2004/10
数値属性を用いたエキスパート付き決定木の構築

全眞嬉, 金子法正, 定兼邦彦, 徳山豪

日本ソフトウェア学会第4回データマイニングワークショップ　2004/09
数値データベースに対するエキスパート付き決定木の構築

全眞嬉, 金子法正, 定兼邦彦, 徳山豪

FIT2004報科学技術フォーラム　2004/09
Efficient Algorithms for Approximating a Multi-Dimensional Voxel Terrain by a Unimodal Terrain International-presentation

Danny Z. Chen, Jinhee Chun, Naoki Katoh, Takeshi Tokuyama

10th Computing and Combinatorics Conference(COCOON2004)　2004/08
Efficient Algorithms for Approximating Multi-Dimensional Voxel Terrain by a Unimodal Terrain

全眞嬉, 定兼邦彦, 徳山豪

日本オペレーションズ・リサーチ学会第9回SAOR研究会　2004/05
Linear Time Algorithm for Approximating a Curve by a Single-Peaked Curve International-presentation

Jinhee Chun, Kunihiko Sadakane, Takeshi Tokuyama

International Symposium on Symbolic and Algebraic Computation (ISAAC03)　2003/12
最適ピラミッドを用いた決定木構築

全眞嬉, 定兼邦彦, 徳山豪

電子情報通信学会第6回情報論的学習理論ワークショップ　2003/11
最適ピラミッドを用いたデータ学習

全眞嬉, 定兼邦彦, 徳山豪

FIT2003情報科学技術フォーラム　2003/09
Linear Time Algorithm for Approximating a Curve by a Single-Peaked Curve International-presentation

Jinhee Chun, Kunihiko Sadakane, Takeshi Tokuyama

Japan-Korea Workshop on Algorithm and Computation (WAAC03)　2003/07
高次元ピラミッドを用いた数値属性結合ルールの生成とデータマイニングへの応用

全眞嬉, Danny Z. Chen, 加藤直樹, 徳山豪

電子情報通信学会第14回データ工学ワークショップ(DEWS2003)　2003/03
Higher-dimensional Pyramid construction Problem and Application to Data Mining

Jinhee Chun,Danny, Z. Chen, Naoki Katoh, Takeshi Tokuyama

情報処理学会88回アルゴリズム研究会　2003/01
Efficient Algorithms for Constructing a Pyramid from a Terrain International-presentation

Jinhee Chun, Naoki Katoh, Takeshi Tokuyama

Japan Conference on Discrete and Computational Geometry 2002（JCDCG02)　2002/12
最適ピラミッドを用いた数値属性ルールの提案

全眞嬉, 定兼邦彦, 徳山豪

日本ソフトウェア学会第4回データマイニングワークショップ　2002/09
地形図からの最適ピラミッドの構成アルゴリズム

全眞嬉, 加藤直樹, 徳山豪

FIT2002報科学技術フォーラム　2002/09
How to Reform a Terrain into a Pyramid

Jinhee Chun, Naoki Katoh, Takeshi Tokuyama

情報処理学会86回アルゴリズム研究会　2002/09

Show all Show first 5

Research Projects 16

Construction of Consistent Computational Geometry in Discrete Spaces

Offer Organization: Japan Society for the Promotion of Science

System: Grants-in-Aid for Scientific Research

Category: Grant-in-Aid for Scientific Research (B)

Institution: Kwansei Gakuin University

2020/04/01 - 2025/03/31
Construction of Consistent Computational Geometry in Discrete Spaces

Offer Organization: Japan Society for the Promotion of Science

System: Grants-in-Aid for Scientific Research

Category: Grant-in-Aid for Scientific Research (B)

Institution: Kwansei Gakuin University

2020/04/01 - 2025/03/31
計算幾何を用いた知識発見とデータ解析手法に関する研究

全眞嬉

Offer Organization: 日本学術振興会

System: 科学研究費助成事業

Category: 基盤研究(C)

Institution: 東北大学

2017/04/01 - 2023/03/31

More details Close

本研究の目的は医療データマイニングにおける、特に、医療画像データ検索における、現在の精度限界を打破するための可視化知識抽出モデルの提案である。医療には膨大な属性の数値と膨大な画像かある。ところが、現在の手法では画像にはノイズか多く、また膨大な容量であるため、解明か困難であり、データマイニング適用に期待か寄せられている。本研究では医療の数値と画像を数値集合として幾何学的に扱い、計算幾何学を用いて巨大数値ベースを効率的に処理する新しい最適化アルゴリズムを与え、上記の精度限界を超える精密な手法の提案を行った。本研究ではイメージ検索モデルの構築と、知識抽出への応用を行い、知識抽出手法における精度限界を超える精密な手法の提案か本研究の目的である。本年度は、医療画像データから知識抽出におけるアルゴリズムを見直し、計算幾何手法を用いたアルゴリズムの設計を行い高速化と知識抽出精度の向上を行なった。特に画像切り出しにおける計算幾何手法を用いた最適かつ高速アルゴリズムの再設計を行い、その提案手法を計算機実装実験を行い、提案手法の挙動と精度を調べた。また、画像切り出しアルゴリズムを我々が提案したオンライン学習理論アルゴリズムに応用を行い、画像切り出しにおいてオンライン学習による機械学習理論への応用研究を行い、計算幾何手法を用いた機械学習理論の提案を行った。その提案手法を論文にまとめて国際会議で発表を行い、高い評価を得た。提案手法を機械学習における説明性の重要な課題を本提案手法を用いることで精度を上げるアルゴリズムの設計を行い、実装実験を行い機械学習における提案手法の妥当性と有効性を調べた。本年度は数値データと画像データから知識抽出において説明性を持ち備えたシステムの構築と探索精度向上と高速化の改良を行った。
非決定性を含む動力学的計算幾何学の理論と応用 Competitive

System: Grant-in-Aid for Scientific Research

2015/04 - 2020/03
実社会ビッグデータ利活用のためのデータ統合・解析技術の研究開発 Competitive

System: The Other Research Programs

2014/04 - 2018/03
革新的研究開発推進プログラム（ImPACT）タフ・ロボティクス・チャレンジ Competitive

System: The Other Research Programs

2014/04 - 2018/03
計算幾何を用いたイメージ検索モデルの構築と知識抽出への応用 Competitive

System: Grant-in-Aid for Scientific Research

2013/04 - 2016/03
医療データおける計算幾何を用いた知識発見システム設計 Competitive

System: Grant-in-Aid for Scientific Research

2011/04 - 2013/03
デジタル空間における計算幾何学の構築とその応用 Competitive

System: Grant-in-Aid for Scientific Research

2010/04 - 2013/03
幾何学データからの物体検知：幾何的なマッチング・切出し・被覆の理論研究 Competitive

System: The Other Research Programs

2010/07 - 2012/06
階層的セグメンテーションルールを用いた数理モデルの構築と知識抽出への応用 Competitive

System: Grant-in-Aid for Scientific Research

2009/04 - 2011/03
大域的ノルムを用いた幾何データ処理最適化理論の構築 Competitive

System: Grant-in-Aid for Scientific Research

2006/04 - 2009/03
数値データに対する階層化ルールと学習理論のデータマイニングへの応用 Competitive

System: Grant-in-Aid for Scientific Research

2007/04 -
計算理論的設計による知識抽出モデルに関する研究 Competitive

System: Grant-in-Aid for Scientific Research

2006/04 -
階層的セグメンテーションを用いたデータマイニングシステムの構築 Competitive

System: Grant-in-Aid for Scientific Research

2005/04 - 2006/03
ファジーセグメンテーションを用いたデータマイニングシステムの構築 Competitive

System: SCF System for Establishment and Support of Center's of Excellence

2003/04 - 2004/03

Show all Show first 5