The growth of mobile services on mobile phones or PDA has been significant in the recent years. This has opened up a new arena for many organizations to implement services on these mobile devices. In this paper, we propose an extension to our work in Artificial Life; of having our embodied conversational agents (ECAs) framework to be extended for mobile handheld devices. We present an overview and investigate new challenges on the implementation of the ECAs on the cross-platform architecture suitable for small end user devices. We demonstrated the proposed downsizing framework suitable for mobile devices with the application in the area of crisis communication.

抄録全体を表示

要旨 2019 年7 月10 日に愛知県新城市の水路において，スジボソハシリグモDolomedes augusti- virgatus によるカワムツNipponocypris temminckii の捕食を観察した．この水路はコンクリート三面張りで，土砂が溜まり植物が生い茂っていた．捕食の様子は，スジボソハシリグモの上顎がカワムツの頭部に深く食い込み，陸上に引きずり上げていた．

抄録全体を表示

2016 年5 月2 日18:13–19:30 に，愛知県豊川市の源流河川でトウカイナガレホトケドジョウ（Lefua sp.2）の追尾行動を確認した．この間に，このドジョウは，大礫と大礫の隙間に入って，それぞれの臀鰭を合わせた．そして，木の枝や大礫の上に乗った卵が確認された．この行動は本種の産卵行動であると思われるため報告する．

抄録全体を表示

This paper describes that a bedding material of carbon fiber has successfully improved human's health at the range that the subject is not aware of own variation in a few days but the precise data showed improvement. For the state of health, there is little accuracy in the conscious answer of a human for a question. In order to solve this problem, a detecting method of physiological body condition for the unconscious responses for the stimuli, reflecting activities of autonomic nervous system during sleep, was tried. Such method was performed by body motion wave (BMW) named by Okawai et al[1]. A bedding material is not only for rest but also, as it were, for rehabilitation tool to recover the state of health.

抄録全体を表示

小児は肺炎への感染や重症化のリスクが高いため，正確な診断や治療が早急に求められる．そこで本論文では，小児肺炎などの胸部疾患を対象としたコンピューター支援診断システムの開発を目的に，解析領域である肺野の抽出手法を提案する．肺野の抽出には，ディープラーニングの一種で，物体検出処理と領域抽出処理を内包しているMask R-CNNを用いた．Mask R-CNNの学習には，ChestX-ray8データベースより選択した小児200枚，成人800枚の合計1000枚の胸部X線画像を用いた．肺炎と診断された小児の胸部X線画像を用いて検証したところ，Jaccard indexの平均値は93.3％，Dice indexの平均値は96.5％となり，高い肺野抽出精度が確認された．

抄録全体を表示

愛知県東三河地方の豊川におけるエビ類の報告は少ない．このため，環境保全活動を行う基礎資料が必要と考えた．そこで2010 年から2013 年までに，豊川でエビ類相の調査を行った．調査を行った場所は豊川の下流域と中流域で，ここに13 地点を設けた．採集道具は，タモ網である．その結果， 7 科16 種のエビ類が確認され，個体数が多く確認されたエビ類はアキアミで，広域で確認されたエビ類はテナガエビであった．確認されたエビ類のなかで明らかな外来種は，アメリカザリガニのみであった．生息環境に対応したエビ類の分布の傾向を把握するために，クラスター分析を行った．汽水域下部には海水種，汽水域上部と淡水域下部には両側回遊種，淡水域中上部には陸封種の分布が関連するという3 つのグループに分けられた．

抄録全体を表示

The synchronization is very important technology in the field of Cellular Neural Networks (CNNs) due to its various applications from the biological, environmental to communication points of view. This paper deals with the parametric synchronization phenomena in large-scaled oscillating CNN. When we determine adequate parameters for the amplitude and frequency in the large-scaled oscillating CNNs with non-oscillating cells, the CNN oscillates in various ways depending on their conditions. In this paper, we have exhibited that both multi-amplitude and multi-phase synchronization are generated by their parameters. To demonstrate them, we simulated multi-dimensional CNN of 4 × 4 (16cells), 8 × 8 (64cells) and 16 × 16 (256cells) and checked the synchronization phenomena. Each cell has an inductor L, a capacitorC and a complex nonlinear conductance as i=f(v). We find that the synchronization phenomena are similarly occurred from the small networks to the larger one. We give a basic theoretical viewpoint for interesting numerical experiment results for the synchronization phenomena in this paper.

抄録全体を表示

We propose a design optimization flow for a high-speed and low-power operational transconductance amplifier (OTA) using a g_m/I_D lookup table design methodology in scaled CMOS. This methodology advantages from using g_m/I_D as a primary design parameter to consider all operation regions including strong, moderate, and weak inversion regions, and enables the lowest power design. SPICE-based lookup table approach is employed to optimize the operation region specified by the g_m/I_D with sufficient accuracy for short-channel transistors. The optimized design flow features 1) a proposal of the worst-case design scenario for specification and g_m/I_D lookup table generations from worst-case SPICE simulations, 2) an optimization procedure accomplished by the combination of analytical and simulation-based approaches in order to eliminate tweaking of circuit parameters, and 3) an additional use of g_m/I_D subplots to take second-order effects into account. A gain-boosted folded-cascode OTA for a switched capacitor circuit is adopted as a target topology to explore the effectiveness of the proposed design methodology for a circuit with complex topology. Analytical expressions of the gain-boosted folded-cascode OTA in terms of DC gain, frequency response and output noise are presented, and detailed optimization of g_m/I_Ds as well as circuit parameters are illustrated. The optimization flow is verified for the application to a residue amplifier in a 10-bit 125MS/s pipeline A/D converter implemented in a 0.18µm CMOS technology. The optimized circuit satisfies the required specification for all corner simulations without additional tweaking of circuit parameters. We finally explore the possibility of applying this design methodology as a technology migration tool, and illustrate the failure analysis by comparing the differences in the g_m/I_D characteristics.

抄録全体を表示

A frontal and realistic face image was synthesized from a single profile face image. It has a wide range of applications in face recognition. Although the frontal face method based on deep learning has made substantial progress in recent years, there is still no guarantee that the generated face has identity consistency and illumination consistency in a significant posture. This paper proposes a novel pixel-based feature regression generative adversarial network (PFR-GAN), which can learn to recover local high-frequency details and preserve identity and illumination frontal face images in an uncontrolled environment. We first propose a Reslu block to obtain richer feature representation and improve the convergence speed of training. We then introduce a feature conversion module to reduce the artifacts caused by face rotation discrepancy, enhance image generation quality, and preserve more high-frequency details of the profile image. We also construct a 30,000 face pose dataset to learn about various uncontrolled field environments. Our dataset includes ages of different races and wild backgrounds, allowing us to handle other datasets and obtain better results. Finally, we introduce a discriminator used for recovering the facial structure of the frontal face images. Quantitative and qualitative experimental results show our PFR-GAN can generate high-quality and high-fidelity frontal face images, and our results are better than the state-of-art results.

抄録全体を表示

An ensemble of multilayer scintillators is discussed as an option of the high-sensitivity detector MOON (Mo Observatory of Neutrinos) for spectroscopic measurements of neutrinoless double beta decays. A prototype detector MOON-1, which consists of 6-layer plastic scintillator plates, was built to study the photon responses of the MOON-type detector. The photon responses, i.e., the number of scintillation photons collected and the energy resolution, which are key elements for high-sensitivity experiments, are found to be 1835±30 photoelectrons for 976 keV electrons and σ=2.9±0.1% (ΔE⁄E=6.8±0.3% in FWHM) at the Q_ββ∼3 MeV region, respectively. The multilayer plastic scintillator structure with high energy resolution as well as a good signal for the background suppression of β–γ rays is crucial for the MOON-type detector to achieve inverted-hierarchy neutrino-mass sensitivity. It will also be useful for medical and other rare-decay experiments as well.

抄録全体を表示

In the summer of 1965,the alpine vegetations of herbaceous and dwarf shrub communities were investigated phytosociologically near the summits of the lide mountains, which are situated only about 40km distant from and facing with no obstacle the Sea of Japan. From the results obtained, the following associations are recognized in the region. 1. Association Primulo-Faurietum crista-galli nov .(Table 1) Ia. Subass, typicum Ib. Subass. veratretosum Ic. Subass. vaccinetosum 2. Association Leontopodio-Arcterietum prov. (Table 3) 2a. Subass. empetretosum 2b. Subass. potentilletosum The two new associations proposed here are obvious to be included in the alliances occurring in the alpine zone, in spite of their having several subalpine components cited in the tables. The occurrence of the alpine communities near the summits of these mountains could be due mostly to the following environmental factor : The area for the associations are covered by deep snow every year. The alpine communities discussed here might have possibly been substitutional vegetations for the coniferous forests. Considering these, the vegetational zone called"pseudo-alpine zone", might be discussed more in the future, because the communities proposed here are considered to be one of the compensational vegetations.

抄録全体を表示

意志性と主語有生性はイベントの基本的な属性であり，密接な関係にある．これらの認識は文脈を考慮したテキスト理解を必要とし，その学習には大量のラベル付きデータを要する．本論文では，人手でラベル付きデータを構築することなく，意志性と主語有生性を同時学習する手法を提案する．提案手法では生コーパス中のイベントにヒューリスティクスを用いてラベルを付与する．意志性のラベルは「わざと」や「うっかり」といった意志性を示す副詞を頼りに付与する．主語有生性のラベルは知識ベースに登録されている有生名詞・無生名詞を頼りに付与する．こうして集めたイベントから手がかり語を含まないイベントに汎化する分類器を構築する．本研究ではこの問題をバイアス削減ないしは教師なしドメイン適応の問題とみなして解く．日本語と英語の実験で，提案手法により，人手でラベル付きデータを構築することなく，意志性・主語有生性の高精度な分類器を構築できることを示した．

抄録全体を表示

Bibliobattle, which is a framework that supports information sharing by introducing books together among participants, has been spread all over Japan. We consider that the Bibliobattle possesses a filtering method that selects a recommendation book from enormous books by presenters. In this paper, we verify what kind of recommendation methods the Bibliobattle possesses when we regard it as a kind of recommender systems. Compared with baseline systems, including a keyword search system, a recommender system by content-based filtering, and that by collaborative filtering, the recommender system by the Bibliobattle is superior to them from the viewpoint of diversity, novelty, and overall satisfaction. In addition, this paper showed that is was superior to them from the viewpoint of explanation because of its characteristics of that the presenters introduce their favorite books together.

抄録全体を表示

本研究では，基本イベントに基づく常識推論データセットの構築手法を提案する．具体的には，テキストから「お腹が空いたので，ご飯を食べる」といった蓋然的関係を持つ基本的なイベント表現の組を自動抽出し，クラウドソーシングによる確認を行った後，基本イベント間の蓋然的関係を問う多肢選択式問題を自動生成する．提案手法に従って 10 万問規模の常識推論データセットを構築し，計算機による解答実験を行った結果，高性能な汎用言語モデルと人間の間に解答精度の開きがあることを示した．また，提案手法の拡張性の高さを利用して疑似問題を大規模に自動生成し，このデータ拡張による常識推論タスクおよび関連タスクでの効果を検証した．実験の結果，蓋然的関係に関する知識を広範に学習させることで，常識推論タスクおよび関連タスクにおいて一定の効果があることを示した．これは，蓋然的関係を推論する能力が自然言語理解において重要であることを示唆している．

抄録全体を表示