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

Research on rapid and accurate image restoration by fusion of signal processing and deep learning

Research Project

Project/Area Number 20K04472
Research Category

Grant-in-Aid for Scientific Research (C)

Allocation TypeMulti-year Fund
Section一般
Review Section Basic Section 21020:Communication and network engineering-related
Research InstitutionKeio University

Principal Investigator

Ikehara Masaaki  慶應義塾大学, 理工学部(矢上), 教授 (00212796)

Project Period (FY) 2020-04-01 – 2023-03-31
Project Status Completed (Fiscal Year 2022)
Budget Amount *help
¥4,290,000 (Direct Cost: ¥3,300,000、Indirect Cost: ¥990,000)
Fiscal Year 2022: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2021: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2020: ¥2,340,000 (Direct Cost: ¥1,800,000、Indirect Cost: ¥540,000)
Keywords画像復元 / 深層学習 / 信号処理 / ボケ除去 / デブラー / DCT / GAN / CNN / 画像処理 / ディフェンシング
Outline of Research at the Start

本研究では、畳み込み層の役割を解明し、信号処理技術により目的に応じた画像の特徴を前処理として検出し、劣化画像を考慮しながら深層学習を行うことにより、より層数の少ないネット構造を用いて、高精度な画像復元技術を開発する。本研究によりカメラ撮影における劣化を低減させるのみならず、画像をセンサーとして用いるセキュリティー分野や社会のあらゆる領域を変革するブレークスルーになり得る。

Outline of Final Research Achievements

In this research project, we developed a high-speed, high-precision image restoration technique by combining signal processing and deep learning. By using signal processing to detect image features as a pre-processing step and deep learning while taking into account the characteristics of degraded images, we have achieved better performance than conventional methods for 1) fog removal, 2) raindrop removal, 3) video super-resolution, and 4) blur removal. Moreover, for convolution, which is a fundamental process in deep learning, we were able to construct a high-performance network while reducing the amount of operations by preparing several kernels and expressing the convolution by linear combination of the kernels.

Academic Significance and Societal Importance of the Research Achievements

本研究では、畳み込み層の役割を解明し、信号処理技術により目的に応じた画像の特徴を前処理として検出し、劣化画像を考慮しながら深層学習を行うことにより、より層数の少ないネット構造を用いて、高精度な画像復元技術を開発する。本研究によりカメラ撮影における劣化を低減させるのみならず、画像をセンサーとして用いるセキュリティー分野や社会のあらゆる領域を変革するブレークスルーになり得る。

Report

(4 results)
  • 2022 Annual Research Report   Final Research Report ( PDF )
  • 2021 Research-status Report
  • 2020 Research-status Report
  • Research Products

    (25 results)

All 2023 2022 2021 2020

All Journal Article (10 results) (of which Peer Reviewed: 10 results,  Open Access: 10 results) Presentation (15 results) (of which Int'l Joint Research: 15 results)

  • [Journal Article] Inverted Residual Fourier Transformation for Lightweight Single Image Deblurring2023

    • Author(s)
      Yae Shunsuke、Ikehara Masaaki
    • Journal Title

      IEEE Access

      Volume: 11 Pages: 29175-29182

    • DOI

      10.1109/access.2023.3243173

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Principal Components of Neural Convolution Filters2022

    • Author(s)
      Fukuzaki Shota、Ikehara Masaaki
    • Journal Title

      IEEE Access

      Volume: 10 Pages: 104328-104336

    • DOI

      10.1109/access.2022.3210710

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Enhanced Multiscale Attention Network for Single Image Dehazing2022

    • Author(s)
      Imai Wataru、Ikehara Masaaki
    • Journal Title

      IEEE Access

      Volume: 10 Pages: 93626-93635

    • DOI

      10.1109/access.2022.3204026

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Single Image Raindrop Removal Using a Non-Local Operator and Feature Maps in the Frequency Domain2022

    • Author(s)
      Ezumi Shinya、Ikehara Masaaki
    • Journal Title

      IEEE Access

      Volume: 10 Pages: 91976-91983

    • DOI

      10.1109/access.2022.3202888

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] 4K Real Time Image to Image Translation Network With Transformers2022

    • Author(s)
      Shibasaki Kei、Fukuzaki Shota、Ikehara Masaaki
    • Journal Title

      IEEE Access

      Volume: 10 Pages: 73057-73067

    • DOI

      10.1109/access.2022.3189649

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Channel Attention GAN Trained With Enhanced Dataset for Single-Image Shadow Removal2022

    • Author(s)
      Ryo Abiko;Masaaki Ikehara
    • Journal Title

      IEEE ACCESS

      Volume: 10 Pages: 12322-12333

    • DOI

      10.1109/access.2022.3147063

    • Related Report
      2021 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] Local Image Denoising Using RAISR2022

    • Author(s)
      Theingi Zin;Shogo Seta;Yusuke Nakahara;Takuro Yamaguchi;Masaaki Ikehara
    • Journal Title

      IEEE ACCESS

      Volume: 10 Pages: 22420-22428

    • DOI

      10.1109/access.2022.3152219

    • Related Report
      2021 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] GAN-Based Image Deblurring Using DCT Loss With Customized Datasets2021

    • Author(s)
      Hiroki Tomosada;Takahiro Kudo;Takanori Fujisawa;Masaaki Ikehara
    • Journal Title

      IEEE ACCESS

      Volume: 9 Pages: 135224-135233

    • DOI

      10.1109/access.2021.3116194

    • Related Report
      2021 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] Residual Learning of Video Frame Interpolation Using Convolutional LSTM2020

    • Author(s)
      Suzuki Keito、Ikehara Masaaki
    • Journal Title

      IEEE Access

      Volume: 8 Pages: 134185-134193

    • DOI

      10.1109/access.2020.3010846

    • Related Report
      2020 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] Multi-Stage Dense CNN Demosaicking With Downsampling and Re-Indexing Structure2020

    • Author(s)
      Yamaguchi Takuro、Ikehara Masaaki
    • Journal Title

      IEEE Access

      Volume: 8 Pages: 175160-175168

    • DOI

      10.1109/access.2020.3025682

    • Related Report
      2020 Research-status Report
    • Peer Reviewed / Open Access
  • [Presentation] Stereo Matching with Supplementary Boundary Information2023

    • Author(s)
      Kaito Hashimoto; Masaaki Ikehara
    • Organizer
      2023 IEEE International Conference on Consumer Electronics (ICCE)
    • Related Report
      2022 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Single Image Fence Removal Using Fast Fourier Transform2023

    • Author(s)
      Keitaro Kume; Masaaki Ikehara
    • Organizer
      2023 IEEE International Conference on Consumer Electronics (ICCE)
    • Related Report
      2022 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Inverted Residual Fourier Transformation for Lightweight Single Image Deblurring2023

    • Author(s)
      Shunsuke Yae; Masaaki Ikehara
    • Organizer
      2023 IEEE International Conference on Consumer Electronics (ICCE)
    • Related Report
      2022 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Single Image Raindrop Removal Using a Non-Local Operator and Feature Maps in the Frequency Domain2023

    • Author(s)
      Shinya Ezumi; Masaaki Ikehara
    • Organizer
      2022 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC)
    • Related Report
      2022 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Multi-Stage Feature Alignment Network for Video Super-Resolution2023

    • Author(s)
      Keito Suzuki; Masaaki Ikehara
    • Organizer
      2022 IEEE International Conference on Image Processing (ICIP) Year: 2022 | Conference Paper | Publisher: IEEE
    • Related Report
      2022 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Scale Recurrent Network for Single Image Dehazing2022

    • Author(s)
      Wataru Imai;Yosuke Ueki;Masaaki Ikehara
    • Organizer
      2022 IEEE International Conference on Consumer Electronics (ICCE)
    • Related Report
      2021 Research-status Report
    • Int'l Joint Research
  • [Presentation] People Removal Using Edge and Depth Information2022

    • Author(s)
      Shunsuke Yae;Masaaki Ikehara
    • Organizer
      2022 IEEE International Conference on Consumer Electronics (ICCE)
    • Related Report
      2021 Research-status Report
    • Int'l Joint Research
  • [Presentation] Underwater Image Enhancement with Multi-Scale Residual Attention Network2021

    • Author(s)
      Yosuke Ueki;Masaaki Ikehara
    • Organizer
      2021 International Conference on Visual Communications and Image Processing (VCIP)
    • Related Report
      2021 Research-status Report
    • Int'l Joint Research
  • [Presentation] High Reflection Removal Using CNN with Detection and Estimation2021

    • Author(s)
      Isana Funahashi;Naoki Yamashita;Taichi Yoshida;Masaaki Ikehara
    • Organizer
      2021 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC)
    • Related Report
      2021 Research-status Report
    • Int'l Joint Research
  • [Presentation] Residual Learning of Video Frame Interpolation Using Convolutional LSTM2021

    • Author(s)
      Suzuki, Keito; Ikehara, Masaaki
    • Organizer
      ICPR
    • Related Report
      2020 Research-status Report
    • Int'l Joint Research
  • [Presentation] GAN-Based Image Deblurring Using DCT Discriminator2021

    • Author(s)
      Tomosada, Hiroki; Kudo, Takahiro; Fujisawa, Takanori; Ikehara, Masaaki
    • Organizer
      ICPR
    • Related Report
      2020 Research-status Report
    • Int'l Joint Research
  • [Presentation] Weighted Generalization of Dark Channel Prior with Adaptive Color Correction for Defogging2020

    • Author(s)
      Yosuke Ueki;Masaaki Ikehara
    • Organizer
      2020 28th European Signal Processing Conference (EUSIPCO)
    • Related Report
      2020 Research-status Report
    • Int'l Joint Research
  • [Presentation] Shadow Detection and Removal Using GAN2020

    • Author(s)
      Takahiro Nagae;Ryo Abiko;Takuro Yamaguchi;Masaaki Ikehara
    • Organizer
      2020 28th European Signal Processing Conference (EUSIPCO)
    • Related Report
      2020 Research-status Report
    • Int'l Joint Research
  • [Presentation] GAN-Based Rain Noise Removal from Single-Image Considering Rain Composite Models2020

    • Author(s)
      Takuro Matsui;Masaaki Ikehara
    • Organizer
      2020 28th European Signal Processing Conference (EUSIPCO)
    • Related Report
      2020 Research-status Report
    • Int'l Joint Research
  • [Presentation] Rapid and Accurate Local Gaussian Noise Removal2020

    • Author(s)
      Shogo Seta;Yusuke Nakahara;Takuro Yamaguchi;Masaaki Ikehara
    • Organizer
      APSIPA2020
    • Related Report
      2020 Research-status Report
    • Int'l Joint Research

URL: 

Published: 2020-04-28   Modified: 2024-01-30  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi