| 澤邊 太志 | 1451055: M, 1回目発表 | 環境知能学 |
|
title: The improvement of the delivery system in the marginal village by using the information technology.
abstract: Recently, there are many marginal villages in Japan. The marginal village in here means the village that has being depopulation and over 50% of residents are more than 65 years old. In these kinds of environment, there are a lot of problems in their daily life compared with normal towns or villages. My researches focus in the problem of the delivery system between villages, and the village to nearby shops. This presentation shows ideas of an optimization of the delivery system in the marginal village by using an information technology. language of the presentation: Japanese | ||
| 石原 葵 | 1451008: M, 1回目発表 | 光メディアインタフェース |
|
title: Designing Mirror System for Hemispherically Arraying Virtual Cameras
abstract: In order to record the light field, it is required to place a number of cameras in arbitrary positions. However it may not practical in cost and interference because of its size. In this study, we create virtual cameras by mirrors as an alternative for the multiple cameras, and propose a way to design a mirror system to put virtual cameras in arbitrary positions. First, we consider designing of mirror system to put the virtual cameras hemispherically-outwardly in order to record the light field in a certain lighting environment. language of the presentation: Japanese 発表題目: 鏡を用いた仮想カメラの半球面上配置の設計 発表概要: 光線空間を記録するためには，複数のカメラを任意の位置へ配置することが必要となるが，コスト面や物理的な干渉による問題から実用的でない場合がある．本研究では，上述した複数のカメラの代替として，鏡を用いて仮想カメラを生成し，仮想カメラを任意の位置へ配置するための鏡の設計方法を提案する．本研究では，まず，任意の照明環境の光線空間を記録することを目的として，多数の仮想カメラを半球面上に外向きに配置するための鏡の設計について検討する． | ||
| 遠藤 栄典 | 1451022: M, 1回目発表 | 視覚情報メディア |
| �睇堯‘道� | 1451064: M, 1回目発表 | 視覚情報メディア |
|
title: Novel View Synthesis with object removal for a real environment measured by an on-vehicle system
abstract: Novel View Synthesis (NVS) is a method that generates an image from a novel viewpoint. Recently, the demand for NVS that can handle a wide environment measured by an on-vehicle system capable of measuring 3D Geometry and Images around the vehicle and 3D position of the vehicle has increased, aiming applications such as Google Earth or a Driving simulator system. Existing NVS methods target at static scenes without moving objects. However, the on-vehicle systems usually suffer from a lot of moving objects such as oncoming vehicles, proceeding vehicles and pedestrians on the streets. This study proposes a Novel View Synthesis method with object removal from measured data so that it can generate novel view from real environment data. This presentation shows problems of the existing NVS method applied to dynamic scenes with moving objects and our preliminary study of moving object detection for locating moving object removal. language of the presentation: Japanese 発表題目: 車載システムにより計測された実環境中の移動物体を除去した自由視点画像生成 発表概要: 自由視点画像生成とは，入力視点には無い仮想視点からの映像を入力データから生成する手法である．近年，Googl Earthやドライブシミュレータなどのアプリケーションにおいて，カメラ，レーザレンジスキャナ，GPSなどを組み合わせた，車載システムを用いて計測された広域空間を対象とした自由視点画像生成の需要が高まっている．既存の自由視点画像生成手法では，主に移動物体のないシーンを対象としている．そのため，撮影は移動物体が写り込まないよう行うのが一般的であるが，非現実的，非実用的である．本研究では，車載システムにより計測した動的シーンにおいても，移動物体を効果的に除去することで自由視点画像生成が可能な手法の確立を目的とする．本報告では従来の単純な自由視点画像生成手法を動的シーンに対して適用することで予備実験を行い，移動物体を含む自由視点画像生成時の問題点や，移動物体の検出方法について検証した結果を報告する． | ||
| 辻岡 聡 | 1451075: M, 1回目発表 | 知能コミュニケーション |
|
title: English automatic speech recognition for non-native English speakers
abstract: With the recent years of globalization, the English language become the nowdays tool to facilitate communication at many international meetings. Consequently, to apply automatic speech recognition (ASR) techniques for meeting dictations, the ASR system shall have the capabilty to recognize the speech utterances not only from native speakers but also from non-native speakers. However, the accuracy of non-native English ASR is still not good, therefore the improvement of it become essential. In this study, I attempt to apply an efficient pronunciation modeling of non-native speakers in order to improve the accuracy of non-native English ASR. language of the presentation: Japanese 発表題目: 非英語母語話者を考慮した英語音声認識システムの構築 発表概要: 近年のグローバル化に伴い、国際会議では事実上、国際標準語である英語で意思疎通を図らなければならない場面が急増している。このような場面において、国際会議のディクテーションを考えた場合、非英語母語音声認識システムの精度向上は必要不可欠である。この音声認識システムの精度向上には、音響モデル、発音辞書、言語モデルにおいて検討する必要があるが、本研究では発音辞書に焦点を当て、非英語母語音声認識システムの精度向上を目指す。 | ||
| 三浦 明波 | 1451101: M, 1回目発表 | 知能コミュニケーション |
|
title: Improvement of pivot translation accuracy by using syntactic information
abstract: Statistical Machine Translation is a technology of the translation of text from one human language to another by a computer that learned how to translate from vast amounts of translated text. However, for some language pairs, large bilingual corpora to produce high quality translation results are not readily available. In this kind of situation, a conventional solution is to introduce a "bridge" (pivot) language to connect the source and target language, and several effective strategies for pivot translation of language pairs having relatively similar structures have been proposed. In this study, we research what pivot translation method is effective for language pairs having problems of resource poorness and structural dissimilarity, such as Japanese and French, and aim to improve the pivot translation accuracy by using syntactic information. language of the presentation: Japanese 発表題目: 統語情報を用いたピボット翻訳精度の向上 発表概要: 統計的機械翻訳では，対訳コーパスから翻訳規則を自動的に獲得し，言語間の翻訳が可能となる．しかし，十分な翻訳精度を得るための大規模コーパスを直ちに取得できないような言語対も数多く存在する．こういった状況では，中間言語を介して翻訳を行うピボット翻訳が有効な解法の一つであり，比較的構造の似通った言語対で有効とされるピボット翻訳手法が既に複数考案されている．そこで本研究では，日本語とフランス語などのように，対訳コーパスを得にくく，かつ言語構造も大きく異なる言語対において，どのようなピボット翻訳手法が有効であるか調査を行い，統語情報を用いることでピボット翻訳の精度向上を目指す． | ||
| 矢野 完人 | 1451114: M, 1回目発表 | 計算システムズ生物学 |
|
title: Three dimensional structural analysis of dendritic spines abstract: Recently, research about the brain have been actively performed. However, we don't know the brain very well. So we need to verify about the brain very well. There are a variety of techniques to analyze the brain. We have analyzed the brains by using two-photon excited fluorescence microscopy. And, I have focused on dendritic spines. I have studied a method of analyzation in the morphological features dendritic spines. language of the presentation: Japanese | ||
| 町田 宗丈 | 1451097: M, 1回目発表 | 数理情報学 |
|
title: Application of fictive-learning model to decision-making model of Obsessive-Compulsive Disorder
abstract:Humans decision-making model receiving attention as solve at mathematically and scientifically to aberration base of various psychiatric disorders. We research aim at apply to treatment obtain relationship decision-making model from analyze behavior data at psychiatric patient and healthy control, and neurotransmitter in Obsessive-Compulsive Disorder. This presentation propose of fictive-learning model to decision-making, and shows result of apply to actual behavior data. language of the presentation:Japanese 発表題目: 強迫性障害患者の行動モデルへのfictive-learningモデルの適用 発表概要:ヒトの意思決定モデルは，さまざまな精神疾患の根本にある異常を数理的および科学的に解明する手段として近年注目されている． 我々は強迫性障害に注目し，精神疾患患者と健常者の行動データから意思決定モデルを解析し，神経伝達物質との関係を解明することで治療に役立てることを目指している． 本発表ではfictive-learningモデルの適用を提案し，実際の行動データに当てはめた結果について述べる． | ||
| 小松 展久 | 1451048: M, 1回目発表 | 大規模システム管理 |
|
title: On evacuation guiding scheme using trajectories of mobile terminals
abstract: When large-scale disasters occur, evacuees have to evacuate to safe places quickly. Our laboratory aims to achieve an evacuation guiding system using mobile terminals of evacuees. Our proposed system suppresses evacuees' operation of mobile terminals as much as possible because they are under stressful and/or dangerous situation during their evacuation. In this presentation, we introduce our proposed scheme and its fundamental characteristics through simulation experiments. language of the presentation: Japanese 発表題目: モバイル端末の軌跡情報を用いた避難誘導方式に関する一検討 発表概要: 大規模災害発生後,被災者は迅速に安全な避難先へと移動する必要がある．発表者の研究室では，被災地における避難者に対し，各自が保有するモバイル端末を介した避難誘導の実現を目指している．その際，災害発生直後の避難者の緊張状態と被災地の危険性を考慮し，避難者のモバイル端末に対する操作を出来る限り抑えることが要求される．本発表では，自動的に取得可能なモバイル端末の軌跡情報を用いた避難誘導方式の提案とその基本的特性に関する評価を行う． | ||
| 朝倉 高也 | 1451003: M, 1回目発表 | 知能システム制御 |
|
title: Optimal Control Synthesis Method with KL Control for Real-time Robot Control abstract: Recently, there are many research about robotic motion control for handling complicated tasks, and optimal control theory is especially expected to accomplish the tasks which consider surrounding environment. Optimal control theory can generate global optimal motion control policy which minimize cost function that shows given task's archivement. However, its calculation time cost is large and not suitable for real-time control of robots. In our research, we focus on the approximate synthesis method of optimal control, which calculate approximate solutions of new tasks by using already calculated solutions of tasks. Our proposed method guarantee optimality of synthesized controller, which is not guaranteed previous synthesis method, by using KL control. language of the presentation: japanese | ||
| RODRIGUEZ RAMIREZ JUAN ESTEBAN | 1451126: M, 1回目発表 | 知能システム制御 |
|
title: Dynamic quantizers design under data rate constraints by using Differential Evolution (DE) algorithm
abstract: Quantizers, which convert continuous-valued signals to discrete-valued ones, play a crucial role in the implementation of networked control systems (NCSs). However, the quantization of the signals degrades the performance of the system. Dynamic quantizers are effective methods to overcome this performance degradation but their design is not an easy task, especially under the presence of data rate constraints in the system. Differential evolution (DE) is a meta-heuristic optimization method that has probed to be useful in the optimization of complex problems even with nonlinear or non-differentiable cost functions. In this study it is considered the design of feedback dynamic quantizers that satisfy data rate constraints by minimizing an error function using DE. The effectiveness of the proposed method is verified by means of simulations and its performance is compared with the particle swarm optimization (PSO) method, showing superior results. language of the presentation: English | ||
| 藤野 啓輔 | 1451090: M, 1回目発表 | ソフトウェア工学 |
|
title: Toward Monitoring Bugs-fixing Process after the Releases in Open Source Software abstract: As open source software (OSS) has become an integral part of today's software businesses, many software companies rely on OSS to develop their customer solutions and products. On the other hand, they face various concerns in using OSS. One of the concerning is software quality, because there are many bugs in OSS soon after the new releases. When the software company begins using OSS, they should use the updated higher quality OSS. Otherwise, company developers often have to download new patches, and may fix bugs by themselves. In this study, we propose a bug fixed-curves to monitor the bug convergence process after the release of OSS. | ||
| 藤原 新 | 1451091: M, 1回目発表 | ソフトウェア工学 |
|
title: Bug Report Recommendation for Code Inspection abstract: Large software projects such as Mozilla Firefox and Eclipse own more than ten thousand bug reports that have been reported but left unresolved. To utilize such a great amount of unresolved bug reports and accelerate bug detection and removal, we propose to recommend a programmer a bug report that likely to contain failure descriptions related to a source file being inspected. We employ the vector space model (VSM) to make a relevancy ranking of bug reports to a given source file. The result of an experiment using data of three open source software projects showed that the accuracies of recommendations ranged from 48.99% to 86.46% in terms of the percentage of recommendations that contained relevant bug reports in a top 10 recommended list. language of the presentation: Japanese | ||
| 原 一貴 | 1451085: M, 1回目発表 | ソフトウェア設計学 |
|
title: Where is the debugger for my software-defined network? [Paper introduction]
| ||
| 安川 達朗 | 1451113: M, 1回目発表 | ユビキタスコンピューティングシステム |
|
title: Serendipitous Event Recommender System using SNS abstract: Existing recommender systems have mainly focused on user’s preference. However, in addition to the user's preference, "serendipity" which means the unexpectedness and novelty, has became more important factor for the future recommender system. In this research, we propose a novel event recommender system that can offer serendipity events for each user by utilizing Social Networking Service(SNS). All the target events in our research are obtained from major event mediums, such as Connpass, ATND, and Doorkeeper. 発表題目: SNSを利用したセレンディピティなイベント推薦手法の提案 発表概要: 従来，多くの推薦システムでは，推薦されたシステムがユーザの嗜好に沿うか否かが重視されてきた．しかし，近年ではこれに加えて推薦されたアイテムに意外性があるかが問われている． 本研究ではソーシャルネットワークングサービス(SNS)を利用して，セレンディピティを実現する推薦手法を提案する．また，推薦対象はConnpass, ATND,Doorkeeperなどで扱われている イベントを対象とする． | ||
| 宵 憲治 | 1451119: M, 1回目発表 | ユビキタスコンピューティングシステム |
|
title: A Context Movement Method for Efficient Testing of Context-Aware Systems
abstract: Context-aware systems provide services adapting to changes in the context which consists of ambient factors including temperature and humidity and user factors including their movement and position. Developing reliable context-aware systems requires through tests with possible contexts. However, reproducing various contexts in real world takes a lot of energy and human costs. The aim of this research is to minimize the cost to reproduce various contexts. In this presentation, we will introduce a tool called PathSim which has been developed in our laboratory for finding the device operation sequence for moving between arbitrary two contexts at the lowest cost. Then we will show outline of the proposed testing method for context-aware systems using PathSim. We also show a future plan of this research. language of the presentation: Japanese 発表題目: コンテキストアウェアシステムの効率よいテストのためのコンテキスト遷移手法 発表概要: 気温や湿度といった環境的要因，人の動きや位置のような人的要因からなるコンテキストの変化に適応しながら動作するコンテキストアウェアシステムの開発では，様々なコンテキストを再現しシステムをテストする必要があり，従来のテスト手法ではコンテキスト再現のためのコストが大きくなってしまう．本研究では，コンテキストアウェアシステムのテストにおけるコンテキスト再現のコストを最小化することを目的とする．本発表では，先行研究として，任意のコンテキスト間を最小のコストで移動させるデバイス操作系列導出ツールPathSimを紹介し，PathSimを利用したコンテキストアウェアシステムのテスト手法の概要と今後の研究予定について述べる． | ||