| 花井 直樹 | 1551081: M, 1回目発表 | インターネット工学 |
|
title: Considering Transform Domain-based Approaches for Detecting Anomalies in Massive Internet Traffic
abstract: Detecting Anomalies in Internet traffic is not only essential, but also a challenging task for network administrators. Attackers hide anomalous traffic data in massive normal traffic streams. Therefore, the more network traffic increases, the more difficult it is to detect anomalies. In this presentation, I will refer to a novel method for detecting traffic anomalies in networks, and I will consider a new direction for extending this type of research. language of the presentation: Japanese | ||
| 小野 健一 | 1551028: M, 1回目発表 | ソフトウェア工学 |
|
title: Prediction of the number of code reviewers in OSS development
abstract: Code review is a widely used method for verifying the quality of the code modified by software developers. In open source software development, usually developers request a review by selecting the desired reviewers. However, reviewers do not always do the review because they do not have the obligation. For this reason, code development suffers frequent delays. In this study, we aim to predict the number of reviewers that will check the code in order to prevent delays in the development. To that extent, here we present our research objectives and future plan. language of the presentation: Japanese | ||
| 守谷 一希 | 1551109: 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: Propose of Indoor Localization Based on Distance-illuminance Model and Switch-on Control of Lighting Devices abstract: In our previous study, we proposed an indoor localization method based on illuminance fingerprinting. This previous method, however, takes a non-negligible human labor to construct multiple distance-illuminance models for different lighting devices since it supposes to use all existing lighting devices installed in the target indoor space and to be able to control these devices via home network. To cope with these problems, our method estimates distances from three newly installed controllable lighting devices based on the illuminance at the target point by alternately turning on each of the devices. Then, the proposed method estimates the position of the target point based on trilateration. Merits of the proposed method are two folds. First, it requires only one distance-illuminance model since we newly install three identical lighting devices. Second, it does not need to consider influences by external lighting devices (or sunlight) since the proposed method measures the difference of the illuminance between before and after turning on a lighting device. We conducted experiments in a room in an ordinary home and confirmed that the proposed method could estimate the position of the illuminance sensor within 0.5m error on average. language of the presentation: Japanese 発表題目: 照明装置の順点灯制御と照度-距離モデルに基づく屋内位置推定手法の提案 発表概要: 照度を用いた位置推定の先行研究として，フィンガープリンティング方式に基づく屋内位置推定手法が提案されている．その手法の問題点として，対象空間内の既設照明装置を利用するため照明装置種類ごとの距離−照度モデルが必要であることや，制御可能照明装置以外の光源の影響を考えていないことが挙げられる．この問題点に対し，提案手法では，新規に設置した 3 つの照明装置を順番に点灯させた際の推定対象地点の照度を基にそれぞれの照明装置からの距離を推定する．新規に設置する照明装置だけを考慮すればよいので，複数の距離-照度モデルが必要なく，他の光源の影響を考える必要もない．一般家庭の部屋を想定した対象空間内での実験により，提案モデルと提案手法を用いて平均 0.5m の誤差で位置推定が可能であることを確認した． | ||
| 亀田 友哉 | 1551036: M, 1回目発表 | コンピューティング・アーキテクチャ |
|
title: Development of Simulator for Cellular Neural Network and Evaluation
abstract: We are developing a cellular neural network(CNN) which features low power consumption, high-density integration of electron devices, and high functionality. In order to determine the best architecture for a CNN, we developed a CNN simulator for character reproduction. Currently, our simulator can reproduce four numbers simultaneously. This result indicates the possibility that a real circuit will also reproduce characters. | ||
| 藤林 大毅 | 1551091: M, 1回目発表 | ソフトウェア工学 |
|
title:Study of OSS introduction optimum time by using the users and the development situation of the OS
abstract:The open source software (OSS) is a software that can be introduced at low cost. Companies that incorporate OSS products and services and take advantage interest are expanding. After OSS release, if a number of bug is reported, there is a problem that is costly in maintenance work for company. So it窶冱 important for company to assess the introduction time. In previous study, we had to assess the time introduced from the number of users of OSS. It is focused on the number of users, we do not consider introducing standards and methods of the company. This study reveals the OSS introduction criteria and methods by questionnaire to company. And we also clarify that development status (the number of changes, the number of changed lines, developers number, defect report number, bug fixes number) and the number of users is associated with time introduction. As a result, we aim to present a proper introduction time for companies. language of the presentation: Japanese | ||
| 田口 雄哉 | 1551058: M, 1回目発表 | 自然言語処理学 |
|
title: Abstractive Single Document Summarization by Neural Networks
abstract: Text summarization is the process of reducing a text document in order to create a summary retaining the most important points of the original document. Most of text summarization is based on extractive and compressive methods. While extractive based methods has limitation, there were a few of research of abstractive summarization methods. However, the recent tide of neural networks can open up new possibilities of abstractive summarization. I introduce recent study of abstractive summarization by encoder-decoder model of recurrent neural networks. language of the presentation: Japanese | ||
| 白川 誠 | 1551052: M, 1回目発表 | 視覚情報メディア |
|
title: Fast Computation of Seamless Video Loops abstract: Short looping videos concisely capture the dynamism of natural scenes. Creating seamless loops usually involves maximizing spatiotemporal consistency and applying Poisson blending. This paper take an end-to-end view of the problem and present new techniques that jointly improve loop quality while also significantly reducing processing time. A key idea is to relax the consistency constraints to anticipate the subsequent blending, thereby enabling looping of low frequency content like moving clouds and changing illumination. This paper also analyze the input video to remove an undesired bias toward short loops. The quality gains are demonstrated visually and confirmed quantitatively using a new gradient-domain consistency metric. This paper improve system performance by classifying potentially loopable pixels and optimizing on a coarse grid while retaining finer detail. Together these techniques reduce computation times from tens of minutes to nearly real-time. language of the presentation: Japanese | ||
| 櫻井 俊 | 1551046: M, 1回目発表 | 光メディアインタフェース |
|
title:Regression analysis for the sugar content distribution measurement of fruit
Sugar content estimation of fruit by near-infrared spectroscopy of point measurement in fruit sorting field, sorting has been carried out. It can spectrometry in plane by using the hyperspectral camera, and it is conceivable to apply the sugar content estimation in two dimensions, but it takes time for photographing. In this study, we identified a small number of wavelengths effective to sugar content estimation, and aims to achieve high-speed measurement using a normal camera and a band-pass filter. First, we select a small number of wavelengths effective to sugar content estimated by using a lasso regression based on an image obtained by photographing the strawberry slices in hyperspectral camera, and the BRIX value that is actually measured by the sugar content meter. In the future, we will work to the realization of high-speed measurement by the camera and the band-pass filter. | ||
| 岡島 知也 | 1551024: M, 1回目発表 | 環境知能学 |
|
title: Paper Introduction: May I help you?: Design of Human-like Polite Approaching Behavior
abstract: When should service staff initiate interaction with a visitor? Neither simply-proactive (e.g. talk to everyone in a sight) nor passive (e.g. wait until being talked to) strategies are desired. This paper reports our modeling of polite approaching behavior. In a shopping mall, there are service staff members who politely approach visitors who need help. Our analysis revealed that staff members are sensitive to "intentions" of nearby visitors. That is, when a visitor intends to talk to a staff member and starts to approach, the staff member also walks a few steps toward the visitors in advance to being talked. Further, even when not being approached, staff members exhibit "availability" behavior in the case that a visitor's intention seems uncertain. We modeled these behaviors that are adaptive to pedestrians' intentions, occurred prior to initiation of conversation. The model was implemented into a robot and tested in a real shopping mall. The experiment confirmed that the proposed method is less intrusive to pedestrians, and that our robot successfully initiated interaction with pedestrians. language of the presentation: Japanese | ||
| 笹野 仁 | 1551047: M, 1回目発表 | 知能コミュニケーション |
|
title: The Simulation of Multimodal lexical acquisition
Recently there are many researches about categorization function in brain using neuroimaging.I introduce the previous research that works on categorization using multimodal robot, and propose the method based on the simulation for expanding previous work to the dialogue system. In order to achieve the natural communication between a dialogue system and human, the similar function is required to the dialogue system as well. language of the presentation: Japanese | ||
| 西川 直希 | 1551072: M, 1回目発表 | 環境知能学 |
|
title:Grasp Type Revisited: A Modern Perspective on A Classical Feature for Vision
abstract:The grasp type provides crucial information about human action. However, recognizing the grasp type from unconstrained scenes is challenging because of the large variations in appearance, occlusions and geometric distortions. In this paper, first we present a convolutional neural network to classify functional hand grasp types. Experiments on a public static scene hand data set validate good performance of the presented method. This study shows that the grasp type is a powerful symbolic representation for action understanding, and thus opens new avenues for future research. language of the presentation:Japanese | ||
| 大崎 洋介 | 1551022: M, 1回目発表 | ロボティクス |
|
title: Various device operation using a service robot
abstract: Recently, inexpensive robot has sold widely. It is expected that robots spread to the society. The robot require to operate the machine to increase the ability of itself. In this paper, We purpose a method that robot operates the machine by recognizing it using convolutional neural networks that can recognize the machine and operating unit. language of the presentation: Japanese | ||
| 伊藤 涼介 | 1551013: M, 1回目発表 | 知能システム制御 |
|
title: Position control of mobile robot by using image projection mapping abstract: This study focuses on the position control of mobile robots. In the conventional control problem, the current position of the robot is necessary to determine the control input. On the other hand, this study proposes a novel control method which does not need the position information. The key idea of the proposed method is the projection mapping of binary images. The control input of robots is determined by using the information on binary images projected onto the ground. In this presentation, a simple control problem is formulated, and its optimal solution is derived. language of the presentation: Japanese | ||
| 藤井 博晃 | 1551089: M, 1回目発表 | 数理情報学 |
| 若松 信孝 | 1551120: M, 1回目発表 | 計算システムズ生物学 |
|
title: Research and development of function estimation method of secondary metabolites by the three-dimentional structure similarity abstract: Eating and metabolism are important to live.Metabolites are made from food by metabolism.Metabolites can be classified into two types, which are primary metabolite and secondary metabolite.The primary metabolite is one of the essential thing to live.The secondary metabolite is not essential thing to live.but it is used as medicine. In this study,I will focus on the secondary metabolite.The biological activity of secondary metabolites have been investigated by analytical experiment.There are about 1.06 million types of secondary metabolites that existed in this world. If we want to examine the biological activity of all secondary metabolites by experiment , it will need long time and research cost.If we are able to predict the biological activity of secondary metabolites,it becomes possible to make efficient experiment.So,I propose a new method for predicting the biological activity of secondary metabolites. language of the presentation: Japanese | ||
| 日朝 祐太 | 1551083: M, 1回目発表 | 生体医用画像 |
|
title: Measurement of rib cage motion from X-ray video using constrained intensity-based 2D-3D registration
abstract: By measurement of the rib cage motion, further improvement of the diagnosis of pulmonary function is expected. For example, when the left and right lung moves asymmetrically, diseases such as cystic fibrosis or funnel chest is suspected. Our goal is to measure the three-dimensional dynamics of the rib cage motion from X-ray video and CT image by using the 2D-3D registration, in this study. Thus, we suggest the optimization based on biomechanics. As an initial study, we report the results of adding a restraint based on the anatomical model of the rib. language of the presentation: Japanese | ||