| 三重野 隆史 | 1351102: M, 2回目発表 | 中村 哲, | 松本 裕治, | 戸田 智基, | Graham Neubig, | Sakriani Sakti |
|
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. In this talk, we propose a new evaluation method for considering the delay time and translation accuracy and report on progress. language of the presentation: Japanese | ||||||
| 折戸 靖幸 | 1351029: M, 2回目発表 | 池田 和司, | 中村 哲, | 柴田 智広, | 爲井 智也 | |
|
title: Evaluation of balancing ability based on system identification of subjects during CoP tracking tasks
abstract: In this study, I employed CoP (Center of Pressure) tracking task that subjects track visual target of their CoP. CoP and skeleton joints are recorded by the Kinect and Wii Balance Board while subjects conducting the tasks. I estimated frequency response between the target signal and the CoP and joint torques. I investigated differences in postural control between young and elderly subjects from the analysis of frequency response and coordinated actions of joints. language of the presentation: Japanese 発表題目: モデルパラメータ同定に基づくバランス能力の評価 発表概要: 本研究では，足圧中心(Center of pressure: CoP)の指標追跡タスクを高齢者及び若年者に課し，Kinect及びWii Balance Boardを用いて被験者の全身の関節位置及びCoPの記録を行った．そして，追跡信号とCoP間の伝達関数の推定及び関節位置からの関節トルクの推定を行い，周波数応答及び関節間の協調動作の解析から高齢者と若年者における姿勢制御の違いを調査した． | ||||||
| 澤田 直哉 | 1351054: M, 2回目発表 | 山口 英, | 安本 慶一 , | 藤川 和利, | 門林 雄基, | 櫨山 寛章 |
|
title:
Evaluation of virtual machine and virtual storage technology for implementing "Follow The Sun" type datacenter
abstract: Power consumption on data centres is one of the most critical issues on service providers. Follow-the-sun scenario which is migrating contents to a data centre located in day time area is a possible solution for the problem. In the scenario, data centres can use solar energy 24-hours by global migration. However, global migration have problems on scalability, scheduling and management. In this talk, we present preliminary measurement results of instance and storage migrations on virtual environment. We clarify the problems on follow-the-sun scenario from the results. language of the presentation: Japanese | ||||||
| 鈴木 貴裕 | 1351061: M, 2回目発表 | 安本 慶一, | 伊藤 実, | 荒川 豊, | 玉井 森彦 |
|
title: A method for supporting drivers to improve driving performance through driving data analysis
abstract: Features while driving such as car speed and fuel consumption rates generally vary depending on the driver’s skill, the road (type and congestion), and the vehicle type. In order to optimize the performance while driving to a destination, such as fuel consumption, time, and smallness of car jolting (which depend on drivers purpose), (1) route recommendation based on the driving skills, and (2) real-time advices for improvement while driving are considered promising approaches. In this research, aiming to confirm that different drivers result in different driving features even in the same driving route, we report the results of our driving feature analysis to the dataset we collected for six drivers. language of the presentation: Japanese 発表題目: 自動車走行データ分析による快適ドライブ支援手法の考案 発表概要: 燃費や走行速度などの自動車の走行時の特性は，一般に，走行する道路や運転者によって異なる．目的地までのドライブパフォーマンス（燃費・揺れの 少なさ・所要時間等，搭乗者の目的により異なる）を最大化するためには、(1)運転者毎の運転技能に合わせた経路推薦をできること、(2)運転者のスキル に応じて運転改善のアドバイスを行うことが考えられる．本研究では，運転者毎の走行特性が同一経路においても異なる特徴を持つ事を確認するために， 6人の被験者による走行データを収集し，走行特性の違いを分析した結果を報告する． | |||||
| 麻田 優真 | 1351004: M, 2回目発表 | 井上 美智子, | 伊藤 実, | 米田 友和, | 大和 勇太 |
|
title: An Efficient Silent Self-Stabilizing Algorithm for 1-Maximal Matching in Anonymous Network
abstract: We propose a new self-stabilizing 1-maximal matching algorithm which is silent and works for anonymous networks without a cycle of a length of a multiple of 3 under a central unfair daemon. Let e be the number of edges and let n be the number of nodes in a graph. The time complexity is O(e) moves. Therefore, the complexity is O(n) moves for trees or rings whose length is not a multiple of 3. language of the presentation: Japanese | |||||
| 紅林 修斗 | 1351039: M, 2回目発表 | 中島 康彦, | 伊藤 実, | 姚 駿, | 高前田 伸也 |
|
title:Parallelization of Shortest Path Search on Multi Platforms
abstract:The demand of applications with irregular memory access like graph processing has increased. it is more difficult to parallelize them than applications with regular memory access such as physics simulation and image processing. In presentation, I show parallelization of the shortest path problem which is one of the graph processing and evaluate on multi platforms. language of the presentation:Japanese 発表題目:最短経路探索の並列化と各種プラットホームによる性能比較 発表概要:近年、大規模グラフ処理への需要が高まっている。従来の画像処理や物理シミュレーションといったメモリアクセスパターンが規則的なものとは異なり、不規則なメモリアクセスパターンを持つため内在する並列性を活用することが困難である。本発表ではグラフ処理の1つである最短経路問題を取り上げ、マルチコアＣＰＵ、ＧＰＵおよびアクセラレータといった複数のプラットフォームを対象として並列化・高速化を試みる。 | |||||
| 清水 怜 | 1351056: M, 2回目発表 | 中島 康彦, | 伊藤 実, | 姚 駿, | 高前田 伸也 |
|
title: Implementation and Evaluation of An Accelerator based on Manymemory Network
abstract: In recent years, demand of large-scale calculation such as weather simulation, fluid analysis and image processing. These applications include a large amount of stencil computation are known for having high parallelism. We focused on data parallelism in this stencil calculation, and implimented array accelerator that was distributed small memory. I will present prototype results and a performance evaluation of multiple chip configuration. language of the presentation: Japanese 発表題目: メモリネットワークベースアクセラレータの試作と評価 発表概要: 近年，気象計算，流体解析，画像処理のような大規模計算の需要が高まっている。それらのアプリケーションはステンシル計算と呼ばれる計算を多く含み、並列度が高いことで知られている。我々は，このステンシル計算におけるデータ並列性に着目し、内部に小規模メモリを分散配置したアレイ型アクセラレータの試作を行った。本発表では試作結果および複数チップ構成時の性能評価結果を示す。 | |||||