• Search Research Projects
  • Search Researchers
  • How to Use
  1. Back to previous page

An efficient learning sample generation in small sample problem for deep learning

Research Project

Project/Area Number 18K11495
Research Category

Grant-in-Aid for Scientific Research (C)

Allocation TypeMulti-year Fund
Section一般
Review Section Basic Section 61050:Intelligent robotics-related
Research InstitutionChukyo University

Principal Investigator

Mekada Yoshito  中京大学, 工学部, 教授 (00282377)

Co-Investigator(Kenkyū-buntansha) 村瀬 洋  名古屋大学, 情報学研究科, 教授 (90362293)
道満 恵介  中京大学, 工学部, 准教授 (90645748)
Project Period (FY) 2018-04-01 – 2021-03-31
Project Status Completed (Fiscal Year 2020)
Budget Amount *help
¥4,290,000 (Direct Cost: ¥3,300,000、Indirect Cost: ¥990,000)
Fiscal Year 2020: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2019: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
Fiscal Year 2018: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
Keywords深層学習 / 画像生成 / 物体検出 / 人の知覚 / 医用画像 / 学習データ生成 / データ生成 / パターン認識 / 画像処理
Outline of Final Research Achievements

This research aims to improve the performance of deep learning for image recognition when the number of available training samples is limited. For this purpose, we have developed a method that maximizes the effectiveness of the training sample generation method. In lesion detection from medical images, it is often difficult to collect a sufficient number of samples. Generating images that account for the target object's location (e.g., its position in the organ, we have shown that the detection accuracy can be improved even with a small number of training data ). In addition, when estimating the characteristics of human perceptual functions, we found that large image deformations can degrade the estimation performance.

Academic Significance and Societal Importance of the Research Achievements

深層学習による画像認識や物体検出において,様々な理由から学習サンプルを多く集めることが困難な問題がある.そのような場合に画像合成や画像生成により学習サンプルを増加させる方法が提案されている.その際に,対象となる物体が存在する状況を想定した画像生成をおこなうことで,単純な画像生成手法に比べて認識精度が向上できることを示した.また画像を生成する場合に付与する画像変形の程度は,対象の画像や推定したい事柄によって変化させないといけないことを示した.

Report

(4 results)
  • 2020 Annual Research Report   Final Research Report ( PDF )
  • 2019 Research-status Report
  • 2018 Research-status Report
  • Research Products

    (22 results)

All 2021 2020 2019 2018 Other

All Journal Article (3 results) (of which Int'l Joint Research: 1 results,  Peer Reviewed: 3 results,  Open Access: 2 results) Presentation (17 results) (of which Int'l Joint Research: 5 results,  Invited: 1 results) Book (1 results) Remarks (1 results)

  • [Journal Article] Lesion Image Generation Using Conditional GAN for Metastatic Liver Cancer Detection2021

    • Author(s)
      Ikeda Yusuke、Graduate School of Engineering, Chukyo University, Japan、Doma Keisuke、Mekada Yoshito、Nawano Shigeru
    • Journal Title

      Journal of Image and Graphics

      Volume: 9 Issue: 1 Pages: 27-30

    • DOI

      10.18178/joig.9.1.27-30

    • Related Report
      2020 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Estimation of the Attractiveness of Food Photography Based on Image Features2019

    • Author(s)
      TAKAHASHI Kazuma、HATTORI Tatsumi、DOMAN Keisuke、KAWANISHI Yasutomo、HIRAYAMA Takatsugu、IDE Ichiro、DEGUCHI Daisuke、MURASE Hiroshi
    • Journal Title

      IEICE Transactions on Information and Systems

      Volume: E102.D Issue: 8 Pages: 1590-1593

    • DOI

      10.1587/transinf.2018EDL8219

    • NAID

      130007686427

    • ISSN
      0916-8532, 1745-1361
    • Year and Date
      2019-08-01
    • Related Report
      2019 Research-status Report
    • Peer Reviewed / Open Access / Int'l Joint Research
  • [Journal Article] Metastatic Liver Cancer Detection by CNN Using Artificial Lesion Images2019

    • Author(s)
      小西 孝明, 道満 恵介, 縄野 繁, 目加田 慶人
    • Journal Title

      Medical Imaging Technology

      Volume: 37 Issue: 1 Pages: 46-50

    • DOI

      10.11409/mit.37.46

    • NAID

      130007584631

    • ISSN
      0288-450X, 2185-3193
    • Year and Date
      2019-01-25
    • Related Report
      2018 Research-status Report
    • Peer Reviewed
  • [Presentation] 料理写真の魅力度推定精度向上のための視線停留分布に対するデータ拡張手法の検討2021

    • Author(s)
      宮崎 光明,服部 竜実,道満 恵介,平山 高嗣,川西 康友,井手 一郎,目加田 慶人
    • Organizer
      動的画像処理実利用化ワークショップ2021
    • Related Report
      2020 Annual Research Report
  • [Presentation] 機械学習を用いた肝がん分子標的療法の効果予測の調査2020

    • Author(s)
      池田 裕亮,道満 恵介,西田 直生志,目加田 慶人
    • Organizer
      第18回情報学ワークショップ
    • Related Report
      2020 Annual Research Report
  • [Presentation] 超音波診断動画からの深層学習手法による肝腫瘍の検出と追跡2020

    • Author(s)
      山岸 生弥,道満 恵介,目加田 慶人
    • Organizer
      第18回情報学ワークショップ
    • Related Report
      2020 Annual Research Report
  • [Presentation] 料理写真の魅力度推定において有効な画像特徴量の検討2020

    • Author(s)
      宮崎 光明,道満 恵介,平山 高嗣,川西 康友,井手 一郎,目加田 慶人
    • Organizer
      第18回情報学ワークショップ
    • Related Report
      2020 Annual Research Report
  • [Presentation] A study on liver tumor detection from an ultrasound image using deep learning2020

    • Author(s)
      Takahiro Nakashima,Issei Tsutsumi,Hiroki Takami,Keisuke Doman,Yoshito Mekada,Naoshi Nishida,Masatoshi Kud
    • Organizer
      Proc. of Joint Int. Workshop on Advanced Image Technology 2020
    • Related Report
      2019 Research-status Report
    • Int'l Joint Research
  • [Presentation] Application of data augmentation for accurate attractiveness estimation for food photography2019

    • Author(s)
      Tatsumi Hattori,Keisuke Doman,Ichiro Ide,Yoshito Mekada
    • Organizer
      Proc. of 11th Workshop on Multimedia for Cooking and Eating Activities
    • Related Report
      2019 Research-status Report
    • Int'l Joint Research
  • [Presentation] A study on a data augmentation framework for accurate food attractiveness estimation2019

    • Author(s)
      服部 竜実,道満 恵介,井手 一郎,目加田 慶人
    • Organizer
      電子情報通信学会 魅力工学研究会シンポジウム2019
    • Related Report
      2019 Research-status Report
  • [Presentation] 視線情報を考慮した料理写真の魅力度推定手法に関する検討2019

    • Author(s)
      宮崎 光明,服部 竜実,道満 恵介,平山 高嗣,川西 康友,井手 一郎,目加田 慶人
    • Organizer
      電子情報通信学会 魅力工学研究会シンポジウム2019
    • Related Report
      2019 Research-status Report
  • [Presentation] 深層学習による超音波画像からの肝腫瘍検出に関する初期的検討2019

    • Author(s)
      堤 一晴,中島 崇博,道満 恵介,目加田 慶人,西田 直生志,工藤 正敏
    • Organizer
      第38回日本医用画像工学会大会(JAMIT)
    • Related Report
      2019 Research-status Report
  • [Presentation] 転移性肝がん検出のためのConditional GANによる学習画像生成2019

    • Author(s)
      池田 裕亮,道満 恵介,目加田 慶人,縄野 繁
    • Organizer
      第38回日本医用画像工学会大会(JAMIT)
    • Related Report
      2019 Research-status Report
  • [Presentation] 超音波画像診断のための深層学習を用いた腫瘍判別2019

    • Author(s)
      堤 一晴,道満 恵介,目加田 慶人,西田 直生志,工藤 正敏
    • Organizer
      2019年度日本生体医工学会東海支部学術集会
    • Related Report
      2019 Research-status Report
  • [Presentation] Character recognition of modern Japanese official documents using CNN for imblanced learning data2019

    • Author(s)
      Zongjhe Yang,Keisuke Doman,Masashi Yamada,Yoshito Mekada
    • Organizer
      Int. Workshop on Advanced Image Technology
    • Related Report
      2018 Research-status Report
    • Int'l Joint Research
  • [Presentation] 機械学習を用いた異常検知による経時天体画像からの変光星候補の絞り込み2019

    • Author(s)
      岩野 勇弥,道満 恵介,高妻 真次郎,目加田 慶人
    • Organizer
      動的画像処理実利用化ワークショップ
    • Related Report
      2018 Research-status Report
  • [Presentation] 料理写真の高精度な魅力度推定のためのデータ拡大の検討2019

    • Author(s)
      服部 竜実,道満 恵介,井手 一郎,目加田 慶人
    • Organizer
      動的画像処理実利用化ワークショップ
    • Related Report
      2018 Research-status Report
  • [Presentation] 肝臓がん検出器のための3D-DCGANを用いた学習用画像生成法2019

    • Author(s)
      池田 裕亮,小西 孝明,道満 恵介,縄野 繁,目加田 慶人
    • Organizer
      第37回日本医用画像工学会大会
    • Related Report
      2018 Research-status Report
  • [Presentation] Learning sample generation for detecting liver cancer using 3D-CNN2018

    • Author(s)
      Yoshito Mekada,Keisuke Doman
    • Organizer
      40th Int. Engineering in Medicine and Biology
    • Related Report
      2018 Research-status Report
    • Int'l Joint Research / Invited
  • [Presentation] A study on the factors affecting the attractiveness of food photography2018

    • Author(s)
      Tatsumi Hattori,Keisuke Doman,Ichiro Ide,Yoshito Mekada
    • Organizer
      10th Workshop on Multimedia for Cooking and Eating Activities
    • Related Report
      2018 Research-status Report
    • Int'l Joint Research
  • [Book] Deep Learning in Medical Image Analysis: Challenges and Applications,Gobert Lee and Hiroshi Fujita eds.(分担執筆)2020

    • Author(s)
      Keisuke Doman,Takaaki Konishi,Yoshito Mekada
    • Total Pages
      176
    • Publisher
      Springer
    • Related Report
      2019 Research-status Report
  • [Remarks] 中京大学目加田・道満研究室

    • URL

      https://md.sist.chukyo-u.ac.jp/

    • Related Report
      2018 Research-status Report

URL: 

Published: 2018-04-23   Modified: 2022-01-27  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi