| 泉川 大樹 | 1351008: M, 1回目発表 | インターネット工学 |
|
title: The abstract of technology about Delay Tolerant Network and its application of InterPlanetary Network
abstract: The future Internet is expected to be tolerant of delay and disruption for the various environments and applications. So far, Delay-/Disruption-Tolerant Network (DTN) has been proposed as a communication method to be tolerant of delay and disruption. DTN can achieve the communication based on store-carry-forward manner, not end-to-end basis. This presentation introduces the overview of DTN technologies and the research trend of Interplanetary Networks. language of the presentation: Japanese 発表題目: Delay Tolerant Network 技術の概要と惑星間通信への応用 発表概要: これからのインターネットは，通信環境およびアプリケーションの多様性により，遅延や切断に対する耐性が求められる．これまで，そのような遅延や切断を許容可能な通信方式として，Delay-/Disruption-Tolerant Network (DTN)が提案されている．DTN は，これまでのエンドツーエンド通信とは異なり，蓄積運搬転送方式による通信を実現する．本発表では，DTN 技術の概要と，DTN を用いた惑星間通信の現在の研究動向について紹介する． | ||
| 澤田 直哉 | 1351054: M, 1回目発表 | インターネット工学 |
|
title: *** Proposal of user-oriented QoS system in a conference events ***
abstract: *** Recent years, there are a lot of demos using smartphone or tablet in conference event. So, there is a need to provide a wireless LAN. But actually, it causes some problem by to much wireless network traffic. Because of it, providing wireless LAN is more difficult than before when there are no smartphones and tablet. Therefore, I propose user based QoS in conference event. *** language of the presentation: *** Japanese *** 発表題目: *** カンファレンスイベントにおけるユーザー指向のQOSシステムの提案 *** 発表概要: *** 近年、学会でのデモ発表等、イベントを行う際に、スマートフォンやタブレットを使用したデモが増加傾向にあり、その事からカンファレンスイベントで無線LANを提供する事への需要が高まっている。しかし、スマートフォンやタブレット端末の普及に伴う無線帯域を利用する端末の増加及びネットワークトラフィックの増加によって、学会等のカンファレンスイベントでの無線LAN環境の提供が難しくなってきている。そこで、カンファレンスイベント時にユーザー指向のQOSシステムの提案を行う。 *** | ||
| 津田 航 | 1351068: M, 1回目発表 | インターネット工学 |
|
title: Research about detection system of Botnet and spam mail
abstract: In recent years, the dangers of spam mail increase and become a social problem.To address this problem,I introduce a thesis about malware analysis by using DNS Traffic and specify related malware domain. moreover, I will discuss about my future research plan. language of the presentation: Japanese | ||
| 髙木 隆志 | 1351063: M, 1回目発表 | ディペンダブルシステム学 |
|
title: Paper introduction : Power Supply Noise Control in Pseudo Functional Test
abstract: In recent years, power supply noise at the time of a delay test for LSI has become a problem. There is a difference in the noise at the time of delay test and normal operation, and it causes the misconception of faulty circuit. Therefore, it is necessary to adjust the noise of the delay test to the same extent as it is of normal operation. Two techniques to adjust the noise are introduced, preamble cycle and Bit-Flip. language of the presentation: Japanese | ||
| 平部 裕子 | 1351091: 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: TouchAnalyzer : a system for using the touch behaviors as a sensor
abstract: In this study, we propose a system for logging, analyzing and visualizing touch operations of users through every application on Android. ‘Touch’ is an important interaction in operating smartphones. Our system uses ’Touch‘ as a new sensor available on smartphones similar to sensors such as accelerometers or gyroscopes in order to recognize user’s context like emotion and skill of the user. The most important point is that it do not require "root" permission i.e. any modification for obtaining the touch operation. Additionally, our proposed system can recognize 8 kinds of representative touch operations. This presentation explains an implementation of our developed tool, and reports obtained results. language of the presentation: Japanese 発表題目: TouchAnalyzer：タッチ操作のログ分析システム 発表概要: 近年スマートフォンやタブレット端末の普及により，タッチ操作はそれら端末を操作するための重要な手段となっている．本研究では，このタッチ操作の挙動を新たなセンサの1つとして着目した．タッチ操作というセンサを用いることで，ユーザの感情やスキルといったより習得が難しいとされるユーザコンテキストを抽出できる可能性が高い．本研究では，Android端末におけるタッチ操作ログを全てのアプリケーションから横断的に取得し，その操作を分析（分類），可視化するシステムTouchAnalyzerを設計・実装した．本システムの重要な点は，タッチ操作を取得するために特別な改造を行う必要がないということである．（OSの改造やroot権限を取得する必要がない）．さらに，本研究で提案するシステムはユーザが日常的に行う8種類の代表的なタッチ操作を識別することが可能である．本発表では，開発したシステムTouchAnalyzerの実装方法と，本システムを用いて得られた結果について述べる． | ||
| 金子 裕哉 | 1351032: M, 1回目発表 | ネットワークシステム学 |
|
title: Bit Error Rate Analysis of OOK signal in Radio on Optical OOK System
abstract: Previous research proposed a configuration of Radio on Fiber (RoF) link that can transfer radio frequency (RF) signal and optical Ethernet OOK signal simultaneously on one fiber by using OOK signal as light source. However, trade-off between strong superimposing of RF signal and Bit Error Rate (BER) of OOK signal is experimentally revealed. In this presentation, BER and optimum threshold of OOK signal in the proposed method are theoretically analyzed. Moreover allowable superimposing power of RF signal is considered based on the theoretical analysis. language of the presentation: Japanese 発表題目: 光OOK重畳ファイバ無線システムにおけるOOK信号の誤り率解析 発表概要: 既設のファイバ網で使われている10Gイーサネット信号を光源とすることで，専用の光源を用いずに無線信号を伝送する ことを目的としたOOK(On-Off Keying)信号を光源とする光ファイバ無線（RoF: Radio on Fiber）が提案されている． しかし無線信号の大きな重畳がOOK信号の誤り率とトレードオフとなることが実験的に明らかにされている． 本発表では提案手法中のOOK信号を復調するための最適しきい値および誤り率の理論的な解析を行い， 重畳可能な無線信号の強さを検討する． | ||
| 植松 裕介 | 1351011: M, 1回目発表 | 環境知能学 |
|
title: Human action recognition from still image using pose estimation
abstract: *** Towards and abnormal action detected from images taken by surveillance cameras, search support videos on the Web, recognizing technique has been proposed are various human action from the video or still image. In this presentation, we introduce the problems and research on human action recognition of existing, we describe the research policy of the future at the end. *** language of the presentation: Japanese 発表題目: 人体姿勢推定を用いた静止画像からの人の行動認識 発表概要: 監視カメラで撮影した静止画からの異常行動検出や、Web上の動画の検索支援に向けて、静止画もしくは映像から人の行動を認識する手法は様々なものが提案されている。本発表では既存の人の行動認識に関する研究と問題点を紹介し、最後に今後の研究方針について述べる。 | ||
| 皆川 正浩 | 1351103: M, 1回目発表 | 自然言語処理学 |
|
title: Automatic Metaphor Recognition with Statistical Language Processing
abstract: A metaphor is an implicit figurative expression, which is said to be a linguistic phenomenon relevant to the way humans understand the world. In recent years, statistical language processing methods have been employed to recognize and understand metaphors automatically in some languages. In this presentation, I introduce an existing method with its problem and what I am going to do in this field. language of the presentation: Japanese | ||
| 三重野 隆史 | 1351102: M, 1回目発表 | 知能コミュニケーション |
|
title: An evaluation method considering the delay time of simultaneous speech
translation systems
abstract: In speech translation systems, there is a certain amount of delay from the start of speech translation to the output of the translated sentence. Conventional speech translation systems need to wait for the end of a sentence of the input sentence, but simultaneous speech translation systems can start to translate without waiting for the end of a sentence. Therefore, simultaneous speech translation systems have the ability to reduce the delay time. There are various studies with respect to division of sentences in simultaneous speech translation systems, but no evaluation method considering the delay time and translation accuracy has been established, and therefore it is difficult to optimize the system. To remedy this problem, in this study we propose a new evaluation method for considering the delay time and translation accuracy in this study. language of the presentation: Japanese 発表題目: 同時音声翻訳システムにおける遅延時間を考慮した評価方法 発表概要: 音声翻訳には、音声認識から翻訳された文を出力するまでにかかる遅延時間が 存在する。従来の音声翻訳では入力文の文末を待つ必要があったが、近年注目されている 同時音声翻訳では、文末を待たずに翻訳を開始することができ、遅延時間を短縮すること ができる。この同時音声音訳において、文の分割に関しての様々な研究が行われているが、 翻訳精度と遅延時間を同時に考慮した評価方法は確立しておらず、システムの最適化を行うこと が困難である。そこで、本研究では同時音声翻訳システムの最適化を行えるよう、翻訳精 度と遅延時間を同時に考慮した新しい評価方法に関しての提案を行う。 | ||
| 室田 勇騎 | 1351108: M, 1回目発表 | 知能コミュニケーション |
|
title: Music signal separation based on bayesian spectral amplitude estimator with automatic target prior adaptation
abstract: Common methods used for audio signal separation, which were mainly developed for speech enhancement, are nonlinear filtering algorithms such as Wiener filtering and the minimum mean-square error short-time spectral amplitude (MMSE-STSA) estimator. In particular, the MMSE-STSA estimator and its extended algorithms are optimal Bayesian estimators based on the a priori speech statistical model, resulting in a great improvement of sound quality. However, these methods cannot be effectively applied to music signals because it is difficult to deal with nonstationary interference signals. Also, the performance has strong dependence on the selection of the a priori statistical model, which cannot be determined automatically. So in this study, I propose a new approach based on the generalized Bayesian estimator with automatic prior estimation suitable for music signal separation. language of the presentation: Japanese 発表題目: 統計モデル推定を用いた一般化MMSE-STSA推定器に基づく音楽音源分離 発表概要: 音源分離技術, 主に雑音下での目的音強調として用いられる手法に, Wiener filtering や MMSE-STSA 推定器のような非線形フィルタリングがある. 特にMMSE-STSA推定器やその拡張手法は音声の統計モデルに基づいたベイズ推定器であり, 質の良い音声を得ることができる. しかし, これらの手法は非線形雑音に上手く対処することができず, 音楽音源分離などでは十分な効果を発揮することができない. また, 音楽信号では音声とは異なり用いる統計モデルも楽器に強く依存する. しかしこの自動決定をブラインドに行うことはできない. そこで本研究では, 与えられたデータから統計モデルの自動決定を行う一般化MMSE-STSAを用いた音源分離手法を提案する. | ||
| 池田 健 | 1351005: M, 1回目発表 | 計算システムズ生物学 |
|
title: Study on statistical analysis of brain PET images
abstract: When evaluating at inspected the PET image, many changes in cerebral blood flow in the mental illness and early in the neurodegenerative disease, the difference in the rate of correct diagnosis by the experience of the radiologist, the reproducibility, three-dimensional lesions even in the same reading shadow person and difficulty in understanding the spread manner is problematic. A method as an alternative to the inspection and to analyze brain functions statistically well studied and have been applied in clinical settings. the normal database by healthy people is essential in carrying out the statistical analysis. In this presentation, I will introduce of statistical parametric mapping and related research used in the statistical analysis. language of the presentation: Japanese | ||
| 大棚 勇輝 | 1351017: M, 1回目発表 | 計算システムズ生物学 |
|
title: Relationship of efficacy and the 3D structure in the secondary metabolites of plants
abstract: Secondary metabolites of plants have utility such as spices and raw materials of medicine. Secondary metabolites are organic compounds that are not directly involved in the normal growth, development, or reproduction of an organism. However, secondary metabolites are a substance specific to the species. There are a large number of secondary metabolites. We don’t know function of many secondary metabolites. In experiment, it is difficult to detect efficacy and tertiary structure from all of the secondary metabolites. In this study, we aim to predict an unknown efficacy of secondary metabolites. We show that there is the relation between efficacy and the tertiary structure. We evaluate the similarity among the tertiary structures of secondary metabolites and perform graph clustering based on the similarity. We analyze the result of graph clustering and hierarchical clustering using value of the biological activity. In this presentation, we report the current progress. language of the presentation: Japanese 発表題目: 植物の二次代謝産物における立体構造と効能の関係 発表概要: 植物の二次代謝産物は、抗生物質のような医薬品の原料や人に対して効能がある香辛料など様々な用途に及ぶ。二次代謝産物とは生命に必須の呼吸などの一次代謝以外で生産される代謝産物である。しかし、二次代謝産物は生物種に固有の物質であり、膨大な数が存在し、効能未知のものも多い。このような二次代謝産物の立体構造や効能を実験で全て明らかにするのは困難である。そこで、本研究では効能と立体構造の関係を体系的に理解して、効能未知の二次代謝産物の効能を予測することを目的とする。二次代謝物の立体構造の類似度を評価し、類似度によるグラフネットワーククラスタリングを行った結果、二次代謝産物の効能のクラスタリングとの関係をみる。本発表では、その進捗を報告する。 | ||
| JAKOVLJEVIC NEMANJA | 1351120: M, 1回目発表 | ロボティクス |
|
title: Aerial based 3D mapping and localisation
abstract: In recent years, there are many research methods that work on digitalisation of large structures and environment. If the target is scanned from the land, certain parts of the structures remain invisible. To complete the digitalization, it is needed to scan from an aerial vehicle. However, aerial vehicles suffer from both internally and externally generated noise and distortion. Moreover, in case of aerial vehicle, payload and flight time are severely limited. Then I introduce the related methods on 3D modeling and localization from the air, and we consider the minimum setup of lightweight, self-sufficient system that would be able to cope with noise and distortion. language of the presentation: English | ||
| 西野 雄太 | 1351083: M, 1回目発表 | 計算システムズ生物学 |
|
title: Codon analysis of algal genes and its relation to metabolism
abstract: Algae are a diverse group of autotrophic organisms, which uses photosynthesis. We focus on the algal secondary metabolic systems, which produce products for growth and development. Previous studies indicated the relation between codon usage and secondary metabolic systems in A.thaliana, however, the relation in algae is not yet known. We study that using the data of algal genes and metabolic pathways from the database called Bicycle. We introduce the methods to study codon usage bias in plant. language of the presentation: Japanese | ||
| 藤田 貴大 | 1351093: M, 1回目発表 | 知能システム制御 |
|
title: A Study on the Robust Performance of Feedforward Learning Control System.
abstract: Feedforward learning control system is one of a practical implementation of MRAC and it can be stabilized with comparative ease. Although this specification corresponds to the robust stability against a mismatch of transfer functions of plant and estimation model, robust control performance against the uncertainty is not guaranteed because a estimator in MRAC fails to learn the plant accurately. To solve this problem, A finite impulse response (FIR) estimation model is employed along with conventional transfer function estimation model. Hence this proposal estimation model approximates the plant more precisely, improvement of the tracking performance is confirmed in some simulation examples. language of the presentation: Japanese | ||