Filters








225 Hits in 0.96 sec

Beach Changes of Mizushima Island in Fukui Prefecture and their Prediction

Takaaki UDA, Yasuhito NOSHI, Yukiyoshi HOSHIGAMI, Yu-uki TSUJIOKA, Takayoshi NISHIMURA
2011 Journal of Japan Society of Civil Engineers Ser B2 (Coastal Engineering)  
doi:10.2208/kaigan.67.i_541 fatcat:s5tygogxqrb4nbrrhbwkzdlwzy

Attention Branch Network: Learning of Attention Mechanism for Visual Explanation [article]

Hiroshi Fukui, Tsubasa Hirakawa, Takayoshi Yamashita, Hironobu Fujiyoshi
2019 arXiv   pre-print
Visual explanation enables human to understand the decision making of Deep Convolutional Neural Network (CNN), but it is insufficient to contribute the performance improvement. In this paper, we focus on the attention map for visual explanation, which represents high response value as the important region in image recognition. This region significantly improves the performance of CNN by introducing an attention mechanism that focuses on a specific region in an image. In this work, we propose
more » ... ention Branch Network (ABN), which extends the top-down visual explanation model by introducing a branch structure with an attention mechanism. ABN can be applicable to several image recognition tasks by introducing a branch for attention mechanism and is trainable for the visual explanation and image recognition in end-to-end manner. We evaluate ABN on several image recognition tasks such as image classification, fine-grained recognition, and multiple facial attributes recognition. Experimental results show that ABN can outperform the accuracy of baseline models on these image recognition tasks while generating an attention map for visual explanation. Our code is available at https://github.com/machine-perception-robotics-group/attention_branch_network.
arXiv:1812.10025v2 fatcat:gla72cnlargcpnqbx4kiqvjm6y

IMPROVEMENT OF ARTIFICIAL REEF ON HAMAJU COAST IN FUKUI PREFECTURE
福井県浜住海岸の人工リーフの改良策の検討

Takaaki UDA, Takayoshi MATSUI, Tomokazu SENSHU, Yukiyoshi HOSHIGAMI, Hiroki OZAWA, Masumi SERIZAWA
2014 Journal of Japan Society of Civil Engineers Ser B3 (Ocean Engineering)  
IMPROVEMENT OF ARTIFICIAL REEF ON HAMAJU COAST IN FUKUI PREFECTURE Takaaki UDA, Takayoshi MATSUI, Tomokazu SENSHU, Yukiyoshi HOSHIGAMI, Hiroki OZAWA and Masumi SERIZAWA 土木学会論文集B3(海洋開発), Vol. 70, No. 2,  ... 
doi:10.2208/jscejoe.70.i_762 fatcat:7unecnkk5ncefckkpkpqoegj4i

NUMERICAL INVESTIGATION ON IMPROVEMENT OF ARTIFICIAL REEF ON HAMAJU COAST IN FUKUI PREFECTURE
福井県浜住海岸の人工リーフの改良策の検討

Takaaki UDA, Takayoshi MATSUI, Fumiaki OZAKI, Yokiyoshi HOSHIGAMI, Masumi SERIZAWA, Shiho MIYAHARA
2013 Journal of Japan Society of Civil Engineers Ser B3 (Ocean Engineering)  
NUMERICAL INVESTIGATION ON IMPROVEMENT OF ARTIFICIAL REEF ON HAMAJU COAST IN FUKUI PREFECTURE Takaaki UDA, Takayoshi MATSUI, Fumiaki OZAKI, Yokiyoshi HOSHIGAMI, Masumi SERIZAWA and Shiho MIYAHARA  ... 
doi:10.2208/jscejoe.69.i_856 fatcat:jjh25dg6ojazfmaje447q2s2r4

Pedestrian detection based on deep convolutional neural network with ensemble inference network

Hiroshi Fukui, Takayoshi Yamashita, Yuji Yamauchi, Hironobu Fujiyoshi, Hiroshi Murase
2015 2015 IEEE Intelligent Vehicles Symposium (IV)  
Pedestrian detection is an active research topic for driving assistance systems. To install pedestrian detection in a regular vehicle, however, there is a need to reduce its cost and ensure high accuracy. Although many approaches have been developed, vision-based methods of pedestrian detection are best suited to these requirements. In this paper, we propose the methods based on Convolutional Neural Networks (CNN) that achieves high accuracy in various fields. To achieve such generalization,
more » ... CNN-based method introduces Random Dropout and Ensemble Inference Network (EIN) to the training and classification processes, respectively. Random Dropout selects units that have a flexible rate, instead of the fixed rate in conventional Dropout. EIN constructs multiple networks that have different structures in fully connected layers. The proposed methods achieves comparable performance to state-of-the-art methods, even though the structure of the proposed methods are considerably simpler.
doi:10.1109/ivs.2015.7225690 dblp:conf/ivs/FukuiYYFM15 fatcat:zzbqcj5yqnhw3ehhg25aazyeou

Enhancement of GLI1-transcriptional activity by β-catenin in human cancer cells

Osamu Maeda, Masashi Kondo, Takayoshi Fujita, Noriyasu Usami, Takayuki Fukui, Kaoru Shimokata, Takafumi Ando, Hidemi Goto, Yoshitaka Sekido
2006 Oncology Reports  
The Hedgehog (Hh) signaling pathway and the Wnt signaling pathway are known to play important roles in carcinogenesis and the progression of various human malignant tumors. Although a relationship between these two pathways has recently been reported, the mechanism by which ß-catenin, one of the key molecules of the Wnt signaling pathway, influences the Hh pathway has not yet been revealed in detail. To clarify the role of ß-catenin in relation to the Hh signaling pathway, we transfected GLI1
more » ... d ß-catenin expression constructs into human malignant cells, including stomach, colon, and lung cancers, and evaluated the luciferase activity of GLI-responsive reporter constructs. While exogenous GLI1 increased the luciferase activity, exogenous ß-catenin also enhanced the activity under overexpression of GLI1. However, co-transfection with T-cell factor (TCF)-4 or lymphocyte enhancer factor (LEF)-1 did not influence the activity, indicating that the enhancement of ß-catenin in relation to the Hh signaling pathway is not TCF/LEFdependent. Our results suggest that ß-catenin might be involved in the Hh signaling pathway via enhancement of the transcriptional activity of GLI.
doi:10.3892/or.16.1.91 fatcat:ibhszvmgbzh6dlbec24eemg7je

Cost-alleviative Learning for Deep Convolutional Neural Network-based Facial Part Labeling

Takayoshi Yamashita, Takaya Nakamura, Hiroshi Fukui, Yuji Yamauchi, Hironobu Fujiyoshi
2015 IPSJ Transactions on Computer Vision and Applications  
Facial part labeling which is parsing semantic components enables high-level facial image analysis, and contributes greatly to face recognition, expression recognition, animation, and synthesis. In this paper, we propose a cost-alleviative learning method that uses a weighted cost function to improve the performance of certain classes during facial part labeling. As the conventional cost function handles the error in all classes equally, the error in a class with a slightly biased prior
more » ... ity tends not to be propagated. The weighted cost function enables the training coefficient for each class to be adjusted. In addition, the boundaries of each class may be recognized after fewer iterations, which will improve the performance. In facial part labeling, the recognition performance of the eye class can be significantly improved using cost-alleviative learning.
doi:10.2197/ipsjtcva.7.99 fatcat:owjoufencjhn7mjftmema6arzu

Signature of hidden order and evidence for periodicity modification inURu2Si2

Rikiya Yoshida, Yoshiaki Nakamura, Masaki Fukui, Yoshinori Haga, Etsuji Yamamoto, Yoshichika Ōnuki, Mario Okawa, Shik Shin, Masaaki Hirai, Yuji Muraoka, Takayoshi Yokoya
2010 Physical Review B  
The detail of electronic structures near the Fermi level in URu 2 Si 2 has been investigated employing stateof-art laser angle-resolved photoemission spectroscopy. The observation of a narrow dispersive band near the Fermi level in the ordered state as well as its absence in a Rh-substituted sample strongly suggest that the emergence of the narrow band is a clear signature of the hidden-order transition. The temperature dependence of the narrow band, which appears at the onset of the
more » ... r transition, invokes the occurrence of periodicity modification in the ordered state, which is shown for the first time by any spectroscopic probe. We compare our data to other previous studies and discuss possible implications.
doi:10.1103/physrevb.82.205108 fatcat:k3v2spapr5biddrijntv6yafkq

Bibliographic title search for medical cardiological "signs" (eponyms)

Takayoshi Kanie, Atsushi Mizuno, Fujimi Kawai, Takanobu Iri, Takahiro Suzuki, Shinichirou Oshima, Sho Fukui, Raku Son, Takahiro Matsuo
2018 Journal of general and family medicine  
medical sign names representing the same phenomenon reported in different articles MD 1 | Atsushi Mizuno MD 1 | Fujimi Kawai 2 | Takanobu Iri MD 3 | Takahiro Suzuki MD 4 | Shinichirou Oshima MD 4 | Sho Fukui  ... 
doi:10.1002/jgf2.208 pmid:30464873 pmcid:PMC6238235 fatcat:34nsahsb6vbqzdelg2efx4vjde

Training of CNN with Heterogeneous Learning for Multiple Pedestrian Attributes Recognition Using Rarity Rate

Hiroshi FUKUI, Takayoshi YAMASHITA, Yuji YAMAUCHI, Hironobu FUJIYOSHI, Hiroshi MURASE
2018 IEICE transactions on information and systems  
Pedestrian attribute information is important function for an advanced driver assistance system (ADAS). Pedestrian attributes such as body pose, face orientation and open umbrella indicate the intended action or state of the pedestrian. Generally, this information is recognized using independent classifiers for each task. Performing all of these separate tasks is too time-consuming at the testing stage. In addition, the processing time increases with increasing number of tasks. To address this
more » ... roblem, multitask learning or heterogeneous learning is performed to train a single classifier to perform multiple tasks. In particular, heterogeneous learning is able to simultaneously train a classifier to perform regression and recognition tasks, which reduces both training and testing time. However, heterogeneous learning tends to result in a lower accuracy rate for classes with few training samples. In this paper, we propose a method to improve the performance of heterogeneous learning for such classes. We introduce a rarity rate based on the importance and class probability of each task. The appropriate rarity rate is assigned to each training sample. Thus, the samples in a mini-batch for training a deep convolutional neural network are augmented according to this rarity rate to focus on the classes with a few samples. Our heterogeneous learning approach with the rarity rate performs pedestrian attribute recognition better, especially for classes representing few training samples. key words: pedestrian attributes recognition, heterogeneous learning, rarity rate
doi:10.1587/transinf.2017mvp0001 fatcat:sqjc67d4snfypenlyiqlaxh4ai

High-throughput Screening of Small Molecule Inhibitors of the Streptococcus Quorum-sensing Signal Pathway

Seiji Ishii, Kenji Fukui, Satoshi Yokoshima, Kazuo Kumagai, Youko Beniyama, Tetsuya Kodama, Tohru Fukuyama, Takayoshi Okabe, Tetsuo Nagano, Hirotatsu Kojima, Takato Yano
2017 Scientific Reports  
doi:10.1038/s41598-017-03567-2 pmid:28642545 pmcid:PMC5481443 fatcat:xrbnhm35b5bsdpmb6wug6zsipy

Reduced Levels of Chloroplast FtsH Protein in Tobacco Mosaic Virus-Infected Tobacco Leaves Accelerate the Hypersensitive Reaction

Shigemi Seo, Masaji Okamoto, Takayoshi Iwai, Megumi Iwano, Kiichi Fukui, Akira Isogai, Nobuyoshi Nakajima, Yuko Ohashi
2000 The Plant Cell  
The digitized images were normalized to enhance the contrast (Fukui, 1986 ) by using Adobe Photoshop version 3.0 (Adobe Systems, Inc., Mountain View, CA).  ... 
doi:10.2307/3871219 pmid:10852937 pmcid:PMC149093 fatcat:nvonlsh66nczznd3ms5fii2qpa

Ultrahigh-resolution laser photoemission study of URu2Si2 across the hidden-order transition

Rikiya Yoshida, Yoshiaki Nakamura, Masaki Fukui, Yoshinori Haga, Etsuji Yamamoto, Yoshichika Ōnuki, Mario Okawa, Shik Shin, Masaaki Hirai, Yuji Muraoka, Takayoshi Yokoya
2011 Journal of Physics and Chemistry of Solids  
We have studied the electronic structures of URu 2 Si 2 employing ultrahigh-resolution laser angle-resolved photoemission spectroscopy. The change of photoemission spectra is investigated across the hidden-order transition, and the emergence of a narrow band is clearly observed near the Fermi level for both (π, 0) and (π, π) directions. In addition, it is shown that tuning of light's polarization allows the signal of a hole-like dispersive feature to enhance. These observations prove that laser
more » ... angle-resolved photoemission spectroscopy is an effective tool for studying the evolution of electronic structures across the hidden-order transition in URu 2 Si 2 .
doi:10.1016/j.jpcs.2010.10.048 fatcat:vadip5fqxfd35kdexh6d2psurm

加工・性質

Osamu SAHEKI, Kiyoshi NAKAZAWA, Teiji ITOH, Takeshi NISHI, Koh MORIYAMA, Hiroshi NOZAKI, Takeshi OKAMURA, Tsuyoshi OHUCHI, Shoji TERAI, Keizou TAKAHASHI, Shigeru MASUMOTO, Masabumi YASUDA (+15 others)
1964 Tetsu to hagane  
doi:10.2355/tetsutohagane1955.50.4_628 fatcat:zxqeaocbn5aihbvkta6yethmsy

加工・性質

Osamu SAHEKI, Kiyoshi NAKAZAWA, Teiji ITOH, Takeshi NISHI, Koh MORIYAMA, Hiroshi NOZAKI, Takeshi OKAMURA, Tsuyoshi OHUCHI, Shoji TERAI, Keizou TAKAHASHI, Shigeru MASUMOTO, Masabumi YASUDA (+15 others)
1964 Tetsu to hagane  
Fig.5.Relation between remainder of molten metal on the weld center line and total upset distance.
doi:10.2355/tetsutohagane1955.50.4_621 fatcat:ivfqhk5yxvc7vbrb6zfqtkduha
« Previous Showing results 1 — 15 out of 225 results