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

Unknown object detection and its learning

Research Project

Project/Area Number 19K20299
Research Category

Grant-in-Aid for Early-Career Scientists

Allocation TypeMulti-year Fund
Review Section Basic Section 61010:Perceptual information processing-related
Research InstitutionTokyo Denki University (2020-2021)
Keio University (2019)

Principal Investigator

Yuko Ozasa  東京電機大学, システムデザイン工学部, 准教授 (20782098)

Project Period (FY) 2019-04-01 – 2022-03-31
Project Status Completed (Fiscal Year 2021)
Budget Amount *help
¥4,160,000 (Direct Cost: ¥3,200,000、Indirect Cost: ¥960,000)
Fiscal Year 2021: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2020: ¥2,210,000 (Direct Cost: ¥1,700,000、Indirect Cost: ¥510,000)
Fiscal Year 2019: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Keywords未知物体検出 / 未知物体 / 物体検出 / ハイパースペクトル / NMS / コンピュータビジョン / パターン認識 / 機械学習 / 知能ロボット
Outline of Research at the Start

本研究では,必要最低限のカテゴリを学習させた学習モデルを予め用意しておき,未学習カテゴリの物体が検出された際に,その物体を追跡し,収集したデータを用いて新しいカテゴリを追加学習していく画像処理技術の枠組みを提案する.

Outline of Final Research Achievements

We proposed a framework for unknown object learning. When an unknown object is detected, the object is tracked and the training data of the object is automatically collected.

Academic Significance and Societal Importance of the Research Achievements

本研究では,必要最低限のカテゴリを学習させた学習モデルを予め用意しておき,未学習カテゴリの物体が検出された際に,その物体を追跡し,収集したデータを用いて新しいカテゴリを追加学習していく画像処理技術の枠組みを提案した.この技術を物体検出・認識が必要となるシステムに組み込むことで,システムは最小限のハードウェアコストで,自律的に未知物体を学習することができるようになる.

Report

(4 results)
  • 2021 Annual Research Report   Final Research Report ( PDF )
  • 2020 Research-status Report
  • 2019 Research-status Report
  • Research Products

    (13 results)

All 2021 2019

All Journal Article (7 results) (of which Int'l Joint Research: 1 results,  Peer Reviewed: 7 results,  Open Access: 5 results) Presentation (6 results) (of which Int'l Joint Research: 3 results)

  • [Journal Article] Training Data Generation Based on Observation Probability Density for Human Pose Refinement2021

    • Author(s)
      Oniki Kazumasa、Keio University, Kanagawa, Japan、Kikuchi Toshiki、Ozasa Yuko
    • Journal Title

      Journal of Image and Graphics

      Volume: 9 Issue: 2 Pages: 50-54

    • DOI

      10.18178/joig.9.2.50-54

    • Related Report
      2021 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Robust Surface Reconstruction of Plant Leaves from 3D Point Clouds2021

    • Author(s)
      Ando Ryuhei、Ozasa Yuko、Guo Wei
    • Journal Title

      Plant Phenomics

      Volume: 2021 Pages: 1-15

    • DOI

      10.34133/2021/3184185

    • Related Report
      2021 Annual Research Report
    • Peer Reviewed / Open Access / Int'l Joint Research
  • [Journal Article] Evaluation of Self-Attention Approach in Hyperspectral Single-Pixel Classification2021

    • Author(s)
      Junkei Okada, Tomoya Kaichi, Yuko Ozasa
    • Journal Title

      Proc. of IIEEJ International Conference on Image Electronics and Visual Computing

      Volume: 2021 Pages: 1-2

    • Related Report
      2021 Annual Research Report
    • Peer Reviewed
  • [Journal Article] A Hyperspectral Approach For Unsupervised Spoof Detection With Intra-Sample Distribution2021

    • Author(s)
      Kaichi Tomoya、Ozasa Yuko
    • Journal Title

      Proc. of IEEE International Conference on Image Processing

      Volume: 2021 Pages: 839-843

    • DOI

      10.1109/icip42928.2021.9506625

    • Related Report
      2021 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Non-Maximum Suppression for Unknown Class Objects using Image Similarity2021

    • Author(s)
      Yoshiaki Homma, Toshiki Kikuchi, and Yuko Ozasa
    • Journal Title

      Proc. of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2021)

      Volume: 5 Pages: 444-449

    • DOI

      10.5220/0010240304440449

    • Related Report
      2020 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] Single-Shot Multi-light-Direction Searching on Discretized Lighting Space2021

    • Author(s)
      Tomoya Kaichi, Toshiki Kikuchi, and Yuko Ozasa
    • Journal Title

      SN Computer Science

      Volume: 2 Issue: 2 Pages: 1-9

    • DOI

      10.1007/s42979-021-00546-3

    • Related Report
      2020 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] A Model Ensemble Approach for Few-Shot Learning Using Aggregated Classifiers2019

    • Author(s)
      Toshiki Kikuchi, Yuko Ozasa
    • Journal Title

      IIEEJ transactions on image electronics and visual computing

      Volume: 7 Pages: 97-105

    • NAID

      130008012590

    • Related Report
      2019 Research-status Report
    • Peer Reviewed / Open Access
  • [Presentation] Evaluation of Self-Attention Approach in Hyperspectral Single-Pixel Classification2021

    • Author(s)
      Junkei Okada, Tomoya Kaichi, Yuko Ozasa
    • Organizer
      IIEEJ International Conference on Image Electronics and Visual Computing
    • Related Report
      2021 Annual Research Report
    • Int'l Joint Research
  • [Presentation] A Hyperspectral Approach for Unsupervised Spoof Detection with Intra-Sample Distoribution2021

    • Author(s)
      Tomoya Kaichi and Yuko Ozasa
    • Organizer
      IEEE International Conference on Image Processing (ICIP)
    • Related Report
      2021 Annual Research Report
    • Int'l Joint Research
  • [Presentation] ハイパースペクトルカメラによる物体追跡の問題点に関する考察2021

    • Author(s)
      岡田純京,西村仁志,小篠裕子
    • Organizer
      情報処理学会全国大会
    • Related Report
      2021 Annual Research Report
  • [Presentation] 隠れた物体の反射光による物体推定2021

    • Author(s)
      白樫奈々,鶏内朋也,小篠裕子
    • Organizer
      情報処理学会全国大会
    • Related Report
      2021 Annual Research Report
  • [Presentation] Non-Maximum Suppression for Unknown Class Objects using Image Similarity2021

    • Author(s)
      Yoshiaki Homma, Toshiki Kikuchi, and Yuko Ozasa
    • Organizer
      the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2021)
    • Related Report
      2020 Research-status Report
    • Int'l Joint Research
  • [Presentation] 検出領域の特徴を用いた未知クラス物体のためのNon-Maximum Suppression2019

    • Author(s)
      本間 喜明, 菊池俊基, 小篠裕子
    • Organizer
      第22回 画像の認識・理解シンポジウム(MIRU2019)
    • Related Report
      2019 Research-status Report

URL: 

Published: 2019-04-18   Modified: 2023-01-30  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi