Image

Yoshihiro Fukuhara

aka gatheluck

I am currently working at ExaWizards Inc. as a ML/MLOps Engineer and developing the MLOps Platform to accelerate the process from PoC to service to scale ML business. With my experience as both a researcher and an engineer, I am passionate about quickly turning research-stage technologies into services and implementing them in society in a way that delivers value to many people.

I am also a headquarter member of cvpaper.challenge, and leads XCCV research group collaborating with PRMU/CVIM mentorship programs. My research interests are primarily in robust machine learning for developing reliable and real-world ready ML systems.

About Me


Biography

2024.04 - ongoing

Waseda University. Research Fellow.

2021.08 - ongoing

ExaWizards Inc. ML/MLOps Engineer.

2018.04 - 2024.03

Department of Applied Physics, Waseda University, Japan.

Waseda University Leading Doctoral Course Embodiment Informatics

Withdrawal from the Doctoral Program with the Completion of Course Requirements

2020.06 - 2021.07

myProduct Inc. AI Engineer.

2018.10 - 2020.03

SE4 Inc. Research Internship.

2018.02 - 2018.09

Fove Inc. Research Internship.

2016.04 - 2018.03

Department of Applied Physics, Waseda University, Japan.

Master of Engineering

2014.06 - 2018.03

Waseda University, Japan. Resident Assistant (RA).

Waseda International Dormitory (WID) Kodaira

[About RA]

2016.07 - 2016.09

University of California, Los Angeles (UCLA). Visiting Student.

Cross-disciplinary Scholars in Science and Technology (CSST), 2016 Participants

[Alumni List] (Acceptance rate of Japanese Students:0.1%)

Supervisor: Prof. Stefano Soatto

2012.04 - 2016.03

Department of Applied Physics, Waseda University, Japan.

Bachelor of Engineering

Publications


Preprint

Enhancing Inverse Problem Solutions with Accurate Surrogate Simulators and Promising Candidates

Akihiro Fujii, Hideki Tsunashima, Yoshihiro Fukuhara, Koji Shimizu, Satoshi Watanabe

arXiv preprint.

[arXiv]

ATRO: Adversarial Training with a Rejection Option

Masahiro Kato, Zhenghang Cui, Yoshihiro Fukuhara

arXiv preprint.

[arXiv]

What Do Adversarially Robust Models Look At?

Takahiro Itazuri*, Yoshihiro Fukuhara*, Hirokatsu Kataoka, Shigeo Morishima

arXiv preprint. (* indicates equal contribution)

[arXiv]

Journal

Image

Property Analysis of Adversarially Robust Representation

Yoshihiro Fukuhara, Takahiro Itazuri, Hirokatsu Kataoka, Shigeo Morishima

Journal of Japan Society for Precision Engineering (JSPE), 2021, Vol.87, Issue 1, Pages 83-91.

[J-STAGE] [PDF]

International Conference

Image

Detect Fake with Fake: Leveraging Synthetic Data-driven Representation for Synthetic Image Detection

Hina Otake, Yoshihiro Fukuhara, Yoshiki Kubotani, Shigeo Morishima

In ECCV 2024 Trust What You learN (TWYN) Workshop, 2024.

[arXiv]

Image

Scapegoat Generation for Privacy Protection from Deepfake

Gido Kato, Yoshihiro Fukuhara, Mariko Isogawa, Hideki Tsunashima, Hirokatsu Kataoka, Shigeo Morishima

International Conference on Image Processing (ICIP), 2023.

[arXiv]

Image

Neural Density-Distance Fields

Itsuki Ueda, Yoshihiro Fukuhara, Hirokatsu Kataoka, Hiroaki Aizawa, Hidehiko Shishido, Itaru Kitahara

European Conference on Computer Vision (ECCV), 2022.

[arXiv] [Project] [GitHub]

Image

RLTutor: Reinforcement Learning Based Adaptive Tutoring System by Modeling Virtual Student with Fewer Interactions

Yoshiki Kubotani, Yoshihiro Fukuhara, Shigeo Morishima

In IJCAI 2021 Workshop on Artificial Intelligence for Education (AI4EDU), 2021.

[arXiv] [GitHub]

Image

Automatic Paper Summary Generation from Visual and Textual Information

Shintaro Yamamoto*, Yoshihiro Fukuhara*, Ryota Suzuki, Shigeo Morishima, Hirokatsu Kataoka

International Conference on Machine Vision (ICMV), 2018. (* indicates equal contribution)

[arXiv] [Project]

Image

Understanding Fake Faces

Ryota Natsume*, Kazuki Inoue*, Yoshihiro Fukuhara, Shintaro Yamamoto, Shigeo Morishima, Hirokatsu Kataoka

ECCV 2018 Workshop on Brain-Driven Computer Vision (BDCV), 2018. (* indicates equal contribution)

[arXiv]

Image

Retexturing under Self-Occlusion using Hierarchical Markers

Shoki Miyagawa, Yoshihiro Fukuhara, Fumiya Narita, Norihiro Ogata, Shigeo Morishima

ACM Special Interest Group on Computer Graphics and Interactive Techniques (SIGGRAPH), 2017.

[PDF]

Image

Fiber-Dependent Approach for Fast Dynamic Character Animation

Ayano Kaneda,Tsukasa Fukusato,Yoshihiro Fukuhara,Takayuki Nakatsuka, Shigeo Morishima

Conference on Computer Animation and Social Agents (CASA), 2017.

[PDF]

Domestic Conference (Reviewed)

Scapegoat Generation for Privacy Protection from Deepfake

Gido Kato, Yoshihiro Fukuhara, Mariko Isogawa, Hideki Tsunashima, Hirokatsu Kataoka, Shigeo Morishima

Meeting on Image Recognition and Understanding (MIRU), 2023. (Short oral)

NeDDF: Neural Density-Distance Fields

Itsuki Ueda, Yoshihiro Fukuhara, Hirokatsu Kataoka, Hiroki Aizawa, Hidehiko Shishido, Itaru Kitahara

Meeting on Image Recognition and Understanding (MIRU), 2022. (Short oral)

Predicting the Effect of Formula-driven Supervised Learning with Intrinsic Dimension

Yusuke Kondo, Yoshihiro Fukuhara, Shin'ichi Satoh

Meeting on Image Recognition and Understanding (MIRU), 2022. (Short oral)

Automatic Real-Time Generation of Secondary Motion Considering Local Body Structure of Characters

Ayano Kaneda, Tsukasa Fukusato, Yoshihiro Fukuhara, Takayuki Nakatsuka, Shigeo Morishima

Computer Entertainment Developers Conference (CEDEC), 2017.

Acceleration of Corotated Elasticity Simulation by Rotation Interpolation

Yoshihiro Fukuhara, Shunsuke Saito, Fumiya Narita, Shigeo Morishima

VC/GCAD Symposium, 2016.

Domestic Conference (Not Reviewed)

Vision and Language for Automatic Paper Summary Generation

Shintaro Yamamoto, Yoshihiro Fukuhara, Ryota Suzuki, Shigeo Morishima, Hirokatsu Kataoka

Pattern Recognition and Media Understanding (PRMU), 2018.

Automatic Paper Summarization System by Integration of Language and Visual Information

Shintaro Yamamoto*, Yoshihiro Fukuhara*, Ryota Suzuki, Shigeo Morishima, Hirokatsu Kataoka

Symposium on Young Research Association for NLP Studies (YANS), 2018. (* indicates equal contribution)

Sfm-based Automatic Forest Measurement using Handheld Camera

Yoshihiro Fukuhara, Yuko Shirai, Shigeo Morishima

Meeting on Image Recognition and Understanding (MIRU), 2018.

Retexturing under Self-Occlusion using Hierarchical Markers

Shoki Miyagawa, Yoshihiro Fukuhara, Fumiya Narita, Norihiro Ogata, Shigeo Morishima

Meeting on Image Recognition and Understanding (MIRU), 2017.

Hierarchical Markers for Retexturing under Self-Occlusion

Shoki Miyagawa, Yoshihiro Fukuhara, Fumiya Narita, Norihiro Ogata, Shigeo Morishima

National Convention of Information Processing Society of Japan (IPSJ), 2017.

Robust Hierarchical Markers for Virtual Try-On

Shoki Miyagawa, Yoshihiro Fukuhara, Fumiya Narita, Norihiro Ogata, Shigeo Morishima

VC/GCAD Symposium, 2017.

Acceleration of Corotated Elasticity Simulation by Rotation Interpolation

Yoshihiro Fukuhara, Shunsuke Saito, Fumiya Narita, Shigeo Morishima

National Convention of Information Processing Society of Japan (IPSJ), 2016.

Miscellaneous


Award

2022.07

MIRU Interactive Session Award, Meeting on Image Recognition and Understanding (MIRU), 2022.

2022.04

Best Wizard Award 2021 2nd half.

Company award from ExaWizards.

2018.08

Incentive Award, Symposium on Young Research Association for NLP Studies (YANS), 2018.

2017.12

JAXA Special Award, HP Mars Home Planet -Education League JP-.

2015.09

Performance Award, Visual Infromation Processing Camp (VIP), 2015.

Talk

2023.07

CVPR 2023 速報

Tutorial, Meeting on Image Recognition and Understanding (MIRU), 2023.

[Slide]

2022.07

Ascender: Accelerator of Scientific Development and Research

cvpaper.challenge conference summer (CCCS), 2022.

[Slide]

2020.06

Adversarial Examples 研究動向

Invited Talk, Symposium of Sensing via Image Infomation (SSII), 2020.

[SSII 2020]

2019.09

Adversarial Examples 分野の動向 -History and Trends of Dark Arts-

cvpaper.challenge CVPR 2019 網羅的サーベイ報告会 @日立製作所 中央研究所.

[Slide] [Slide(+Meta Survey)]

Book

2024.09

コンピュータビジョン最前線 Autumn 2024

pp.35-56, イマドキノ論文サーベイ.

[amazon]

2023.05

AI白書 2023

pp.59-69, 開発・運用の基盤、手順、MLOps.

[amazon]

2022.09

コンピュータビジョン最前線 Spring 2022

pp.49-64, フカヨミ敵対的サンプル.

[amazon]

Funding

2016.04 - 2020.03

Graduate Program for Embodiment Informatics Scholarship.

2016.07 - 2016.09

University of California, Los Angeles (UCLA), Cross-disciplinary Scholars in Science and Technology (CSST) Scholarship Program 2016.

2012.04 - 2013.03

Waseda University, Okuma Memorial Special Scholarship.

Patent

画像処理装置及び画像処理方法

特開 2018-195038 (P2018-195038A)

[IP Force]

Certification

CG-ARTS CG Engineer Examination Expert.