Faculty Profiles - Deguchi Daisuke

Presentations -

Division display 181 - 200 of about 727 ／ All the affair displays >>

A Study on Similar Geo-region Mining based on Spatio-temporal Clustering of Social Media Photos

Chen Lu, Yasutomo Kawanishi, Ichiro Ide, Takatsugu Hirayama, Keisuke Doman, Daisuke Deguchi, Hiroshi Murase

IEICE Technical Report (MVE) 2018.10.26

Estimation of Driver's Behavior based on Pedestrian's Attributes when Passing by a Pedestrian

Fumito Shinmura, Yasutomo Kawanishi, Daisuke Deguchi, Takatsugu Hirayama, Ichiro Ide, Hiroshi Murase, Hironobu Fujiyoshi

2018.10.17

Scene-wise Reliability Estimation of Pedestrian Detection using LiDAR

Haruya Kyutoku, Yasutomo Kawanishi, Daisuke Deguchi, Ichiro Ide, Kazuki Kato, Hiroshi Murase

2018.10.17

Epipolar geometry-based ego-localization using an in-vehicle monocular camera International conference

Haruya Kyutoku, Yasutomo Kawanishi, Daisuke Deguchi, Ichiro Ide, Hiroshi Murase

Workshop on Digital Signal Processing for In-Vehicle Systems 2018.10.8

Event date： 2018

Language：English Presentation type：Poster presentation

Venue：Nagoya University, Japan

Nowadays, development of driving support systems and autonomous driving systems have become active.
Vehicle ego-localization using in-vehicle sensors is one of the most important technologies for these systems.
Accordingly, various attempts to localize own vehicle from in-vehicle sensors have been made.
In general, the estimation accuracy of the traveling direction is lower than in the lateral direction.
Therefore, we propose a highly accurate method for ego-localization of the traveling direction based on epipolar geometry using an in-vehicle monocular camera.
The proposed method makes correspondences between in-vehicle camera images and database images with location information, and calculates the location using locations annotated to the corresponding database images.
However, there are many gaps due to the difference of speed and trajectory of vehicles even if the images are obtained along the same road.
To overcome this problem, the distance between the input image and the database image is calculated by the distance metric based on the epipolar geometry and the local feature method.
An experiment was conducted using actual images with correct locations, and we confirmed the effectiveness of the proposed method from its results.

Vehicle counting via car parts detection from an in-vehicle camera image International conference

Fumito Shinmura, Yasutomo Kawanishi, Daisuke Deguchi, Ichiro Ide, Hiroshi Murase

Workshop on Digital Signal Processing for In-Vehicle Systems 2018.10.8

Event date： 2018

Language：English Presentation type：Poster presentation

Venue：Nagoya University, Japan

This paper proposes a method to count vehicles from an in-vehicle camera image by regression based on car parts detection. In the case of an in-vehicle camera image, since vehicles are frequently occluded by other vehicles in traffic congestion, it is difficult to accurately count vehicles. Therefore, we propose a method to count vehicles by regression based on the number of visible car parts. For this, we make an estimator by learning the relation between the number of visible car parts and that of vehicles by Support Vector Regression. We evaluated our method using in-vehicle camera images recorded in an actual environment, where the proposed method performed better than counting detected vehicles.

HOYO: A Gait Dataset Annotated with Mimetic Words

Hirotaka Kato, Takatsugu Hirayama, Keisuke Doman, Yasutomo Kawanishi, Ichiro Ide, Daisuke Deguchi, Hiroshi Murase

IEICE Technical Report (MVE) 2018.9.7

地域・時期ごとのSNS投稿写真に基づく類似地域マイニング

陳璐, 川西康友, 井手一郎, 平山高嗣, 道満恵介, 出口大輔, 村瀬洋

電子情報通信学会魅力工学研究会シンポジウム 2018 2018.9.5

視線情報を利用した料理写真の魅力度推定手法

佐藤陽昇, 平山高嗣, 道満恵介, 川西康友, 井手一郎, 出口大輔, 村瀬洋

電子情報通信学会魅力工学研究会シンポジウム 2018 2018.9.5

A Study on the Detection of Users of White Cane Users focusing on Human Pose

Naoki Nishida, Yasutomo Kawanishi, Daisuke Deguchi, Ichiro Ide, Hiroshi Murase, Jun Piao

Tokai-Section Joint Conference on Electrical, Electronics, Information, and Related Engineering 2018 2018.9.4

A Study on the Construction of a Traffic Signal Detector by Automatic Accumulation of Training Data Leveraging Driving Behavior Information

Takuya Migishima, Haruya Kyutoku, Yasutomo Kawanishi, Daisuke Deguchi, Ichiro Ide, Hiroshi Murase

Tokai-Section Joint Conference on Electrical, Electronics, Information, and Related Engineering 2018 2018.9.3

オノマトペによる歩容の記述の高精度化に向けたデータセットの構築

加藤大貴, 平山高嗣, 道満恵介, 川西康友, 井手一郎, 出口大輔, 村瀬洋

画像の認識・理解シンポジウム(MIRU) 2018.8.8

同一経路走行映像群からのネガティブサンプル自動抽出による人物検出器の高精度化

本谷真志, 久徳遙矢, 出口大輔, 川西康友, 井手一郎, 村瀬洋

画像の認識・理解シンポジウム(MIRU) 2018.8.8

On Understanding Visual Relationships of Concepts by Visualizing Bag-of-Visual-Words Models

Marc Aurel Kastner, Ichiro Ide, Yasutomo Kawanishi, Takatsugu Hirayama, Daisuke Deguchi, Hiroshi Murase

Meeting on Image Recognition & Understanding 2018.8.8

Event date： 2018

Language：English Presentation type：Poster presentation

Venue：札幌コンベンションセンター

Recent applications in image processing often use a multimodal approach using both text and imagery. This is prone to semantic gap issues when converting between image and language. There has been few research quantifiying visual differences when assessing semantic relationships. In this research, we analyze datasets composed of logically related concepts. By visualizing a Bag-of-Visual-Words (BoVW) model spatially, visual semantics of logically related sub-concepts are shown. To find hidden semantics of related concepts, the most common visual words of an image in relation to its neighbors are highlighted. This provides additional semantic knowledge on how sub-ordinate concepts visually relate to another. It is thought to give an insight on the human perception of these concepts, and can be used in future research to estimate psycholinguistic ratings.

Active Scan LIDARによる歩行者の向きを考慮した効率的な歩行者スキャン手法

山本大貴, 新村文郷, 出口大輔, 川西康友, 井手一郎, 村瀬洋

画像の認識・理解シンポジウム(MIRU) 2018.8.7

Event date： 2018

Language：Japanese Presentation type：Poster presentation

Venue：札幌コンベンションセンター

近年，LASER 光の照射方向を瞬時かつ任意方向に制御可能な Active Scan LIDAR が開発中であり，局所的にスキャン密度を高めることにより遠距離の歩行者を検出可能になると期待されている．しかし，歩行者を効率的にスキャンする方法は実現されていない．そこで本発表では，歩行者の向き毎に構築した歩行者形状モデルと存在確率に基づいた歩行者尤度推定によるActive Scan LIDAR のための効率的な歩行者スキャン手法を提案する．

車両周辺環境の違いに応じた歩行者検出信頼度の推定

久徳遙矢, 川西康友, 出口大輔, 井手一郎, 加藤一樹, 村瀬洋

画像の認識・理解シンポジウム(MIRU) 2018.8.6

Event date： 2018

Language：Japanese Presentation type：Poster presentation

Venue：札幌コンベンションセンター

車載カメラを用いた歩行者検出結果は，実環境下において依然として改善の余地がある．
そこで我々はこれまでに，ある検出器の出力に未検出や誤検出が含まれない可能性を，検出器の出力とは独立に検出器の信頼度として出力するシステムを提案してきた．
その中で，カメラから得た各画像の走行環境に応じた未検出と誤検出に関する信頼度について定義した．
本稿では，これらを入力画像から直接推定する信頼度推定器を構築し，実走行データを用いて推定精度を評価した

姿勢を表現する多様体に基づくGANsを用いた特定クラス物体姿勢推定の検討

川西康友, 出口大輔, 井手一郎, 村瀬洋

画像の認識・理解シンポジウム(MIRU) 2018.8.6

Event date： 2018

Language：Japanese Presentation type：Poster presentation

Venue：札幌コンベンションセンター

Generative Adversarial Nets（GANs）は，ある事前分布から画像などのデータを生成可能なネットワークであり，乱数から様々なデータを生成可能である．本研究では，物体の姿勢変化は潜在的には多様体で表現可能であることに着目し，多様体上に定義した確率分布から画像を生成するGANs の枠組みと，それを用いて学習画像を補間しつつ姿勢推定器を学習する手法を提案する．実験では，提案手法により，同一クラスに属する，学習データに含まれない物体の姿勢が補間できることを確認し，また，生成した画像を用いて姿勢推定器を学習することにより，高精度な姿勢推定ができることを示す．

Development of NU Server System to Link between Paper Report and LMS

Shunya Seiya, Ryuya Ito, Kosuke Okamoto, Ukyo Tanikawa, Shigeki Ohira, Daisuke Deguchi, Tomoki Toda

2017.12.15

A Preliminary Study on the Detection of Wheelchair Users using Faster R-CNN

Ukyo Tanikawa, Yasutomo Kawanishi, Daisuke Deguchi, Ichiro Ide, Hiroshi Murase

Tokai-Section Joint Conference on Electrical, Electronics, Information, and Related Engineering 2017 2017.9.7

Improvement of an attractiveness estimation method for food photos considering the appearance of main ingredients

Kazuma Takahashi, Keisuke Doman, Yasutomo Kawanishi, Takatsugu Hirayama, Ichiro Ide, Daisuke Deguchi, Hiroshi Murase

IEICE Technical Report (MVE) 2017.3.6

Active Scan LIDARを用いた歩行者検出のための効率的スキャン法

山本大貴, 出口大輔, 川西康友, 井手一郎, 村瀬洋

電子情報通信学会東海支部卒業研究発表会 2017.3.3

Event date： 2017

Language：Japanese Presentation type：Oral presentation (general)

Venue：名古屋大学

遠方の歩行者においても局所的にスキャンの密度を高めることが可能なActive Scan LIDARが開発中である．
本研究では，歩行者の形状や存在確率を用いることで，Active Scan LIDARを用いた歩行者検出のための歩行者に対する効率的なスキャン法を提案する．

　 PREV - NEXT 　

<Deguchi Daisuke>

Presentations -