CV
Education
Graduate School of Waseda University
(Apr 2020 to Present, Japan)Faculty of Science and Engineering
School of Advanced Science and Engineering
Dept. of Physics and Applied PhysicsGraduate School of Kogakuin University
(Apr 2018 to Mar 2020, Japan)School of Engineering
Dept. of Electrical Engineering and ElectronicsKogakuin University
(Apr 2014 to Mar 2018, Japan)Faculty of Engineering
Dept. of Information and Communications Engineering
Work Experience
OMRON SINIC X
Internship
(May 2022 to Oct 2022, Japan)Perception Group
Skillup AI
Supervisor and Teacher
(Dec 2020 to Present, Japan)Supervise the class of Generative Adversarial Networks (GANs) and teach the class of GANs.
AI-SCHOLOR
Writer
(Dec 2019 to Present, Japan)Write articles as for artificial intelligence, such as Deep Learning, Human Computer Interaction, and so on.
National Institute of Advanced Industrial Science and Technology (AIST)
Research Assistant
(June 2020 to May 2022, Japan)Artificial Intelligence Research Center (AIRC)
Computer Vision Research TeamBabel, Inc.
Consultant for iface
(Apr 2021 to Dec 2021, Japan)Advise about the face swapping application (iface) for improving quality.
National Institute of Advanced Industrial Science and Technology (AIST)
Research Assistant
(June 2019 to Mar 2020, Japan)Intelligent Systems Research Institute
Computer Vision Research Group
Grant
キオクシア若手奨励研究 (KIOXIA Young Scientist Research)
(Jul 2023 to Mar 2024)アーリーバードプログラム (Early Bird Program in Waseda University)
(Jun 2023 to Mar 2024)日本学術振興会特別研究員DC2 (Japan Society for the Promotion of Science)
Research Fellowship for Young Scientists
(Apr 2022 to Mar 2024)早稲田オープン・イノベーション・エコシステム挑戦的研究プログラム / 次世代研究者挑戦的研究プログラム (W-SPRING: Waseda University Open Innovation Ecosystem Program for Pioneering Research / SPRING: Support for Pioneering Research Initiated by the Next Generation)
(Oct 2021 to Mar 2022, Japan)
Skills
- Computer Vision
- Deep Generative Models (GANs, VAE, and Diffusion Models)
- Virtual Try-on (2D-based)
- Reinforcement Learning (Offline RL and Online RL)
- Deep Model Compressions (Distillation)
- Object-oriented Representation Learning
- Programming
- Python (main language)
- C
- C++
- Java
- Deep Learning Framework
- PyTorch (main framework)
- Parallelize GPUs (DDP)
- Chainer
- Parallelize GPUs (ChainerMN)
- PyTorch (main framework)
- Math
- Linear Algebra
- Statistics
- Persistent Homology
- Language
- Japanese
- English
Publications
Journal
Kazuya NAKAMURA, Shugo YAMAGUCHI, Hideki TSUNASHIMA, Shigeo MORISHIMA. “Invertible Fingerprint Replacement for Image Privacy Protection”, IIEEJ Transactions on Image Electronics and Visual Computing, 10 (1), 66, June 2022.
Hideki Tsunashima, Kosuke Arase, Antony Lam, and Hirokatsu Kataoka. “UVIRT—Unsupervised Virtual Try-on Using Disentangled Clothing and Person Features”, Sensors, 20 (19), 5647, Oct 2020.
中村充志, 瀧澤生, 星泰成, 綱島秀樹, 陳キュウ. “画像の感性を反映させたフォントの自動生成”, 日本感性工学会論文誌, 日本感性工学会, 17巻5号, p.523-529, 2018.
International Conference
Ryosuke Oshima, Seitaro Shinagawa, Hideki Tsunashima, Qi Feng, and Shigeo Morishima. “Pointing out Human Answer Mistakes in a Goal-Oriented Visual Dialogue”, In proceedings of International Conference on Computer Vision Workshop on Vision-and-Language Algorithmic Reasoning (VLAR), Paris, Oct 2023.
Gido Kato, Yoshihiro Hukuhara, Mariko Isogawa, Hideki Tsunashima, Hirokatsu Kataoka, and Shigeo Morishima. “SCAPEGOAT GENERATION FOR PRIVACY PROTECTION FROM DEEPFAKE”, In proceedings of International Conference on Image Processing (ICIP), Kuala Lumpur, Oct 2023.
Shinei Arakawa, Hideki Tsunashima, Daichi Horita, Keitaro Tanaka, and Shigeo Morishima. “Memory Efficient Diffusion Probabilistic Models via Patch-based Generation”, In proceedings of CVPR Workshop on Generative Models for Computer Vision (GCV), Vancouver, June 2023. (Oral)
Kazuya Nakamura, Shugo Yamaguchi, Hideki Tsunashima, Shigeo Morishima. “INVERTIBLE FINGERPRINT REPLACEMENT FOR IMAGE PRIVACY PROTECTION”, In proceedings of The 7th IIEEJ International Conference on Image Electronics and Visual Computing (IEVC), Online, Sept 2021. (Oral)
Hideki Tsunashima, Hirokatsu Kataoka, Junji Yamato, Qiu Chen, and Shigeo Morishima. “Adversarial Knowledge Distillation for a Compact Generator”, In proceedings of 25th International Conference on Pattern Recognition (ICPR), Pages 10636-10643, Online, Jan 2021. (Poster)
Hideki Tsunashima, Hirokatsu Kataoka, Junji Yamato, Shigeo Morishima and Qiu Chen. “Adversarial Knowledge Distillation Algorithm for a Compact Generator”, In proceedings of ISAT-18, O-I-2, Tainan, Dec 2019. (Oral, Excellent Oral Paper Presentation Award)
Hideki Tsunashima, Taisei Hoshi and Qiu Chen. “DzGAN: Improved Conditional Generative Adversarial Nets Using Divided z-vector”, ACM International Conference on Computing and Big Data (ICCBD), Pages 52-55, Charleston, Sept 2018. (Oral)
Domestic Conference
荒川深映, Erik Härkönen, 綱島秀樹, 堀田大地, 田中啓太郎, 森島繁生. “拡散モデルを用いたパッチ単位の任意スケール画像生成”, Visual Computing (VC), 東京, 2023年9月. (Poster, VCポスター賞)
Gido Kato, Yoshihiro Hukuhara, Mariko Isogawa, Hideki Tsunashima, Hirokatsu Kataoka, and Shigeo Morishima. “SCAPEGOAT GENERATION FOR PRIVACY PROTECTION FROM DEEPFAKE”, MIRU, Hamamatsu, Jul 2023. (Short-oral)
荒川深映, 綱島秀樹, 堀田大地, 田中啓太郎, 森島繁生. “パッチ分割による拡散確率モデルのメモリ消費量削減の検討”, MIRU, 浜松, 2023年7月. (Poster)
大島遼祐, 品川政太朗, 綱島秀樹, 森島繁生. “Visual Dialogueにおける人間の応答ミス指摘の検討”, MIRU, 浜松, 2023年7月. (Poster)
大島遼祐, 品川政太朗, 綱島秀樹, 森島繁生. “視覚情報を用いたタスク指向型対話における人間の応答に対する間違い指摘の検討”, NLP, 沖縄, 2023年3月. (Poster)
大島遼祐, 品川政太朗, 綱島秀樹, 森島繁生. “視覚情報に基づくタスク指向型対話における人間の応答に対する間違い指摘の検討”, IPSJ, 東京, 2023年3月. (Oral, 学生奨励賞)
荒川深映, 綱島秀樹, 堀田大地, 森島繁生. “複数解像度で画像を生成可能な拡散確率モデル”, IPSJ, 東京, 2023年3月. (Oral, 学生奨励賞)
Hideki Tsunashima, Kosuke Arase, Antony Lam, Seito Kasai, Hirokatsu Kataoka. “High Fidelity Virtual Try-on with Self-Supervised Appearance Flow Estimation without Paired Data”, The 25th Meeting on Image Recognition and Understanding (MIRU2022), Hyogo, July 2022. (Short-oral, インタラクティブセッション賞 (Poster Presentation Award))
Hideki Tsunashima, Kosuke Arase, Antony Lam, Seito Kasai, Hirokatsu Kataoka. “Lightning-fast Virtual Try-on without Paired Data and Direct Supervision”, The 24th Meeting on Image Recognition and Understanding (MIRU2021), Online, July 2021. (Long-oral, 学生優秀賞 (Student Best Paper Award))
綱島秀樹, 邱玥, 片岡裕雄, 森島繁生. “Object-oriented Representation Learningの実世界データ適用に向けた最新手法の性能分析”, Visual Computing 2020, P20, Online, Dec 2020. (Poster)
綱島秀樹, 大川武彦, 相澤宏旭, 片岡裕雄, 森島繁生. “Object-aware表現学習の安定化のためのKLダイバージェンスの周期性アニーリング”, MIRU2020, IS3-2-33, Online, Aug 2020. (Poster)
綱島秀樹, 丸山洸太, 片岡裕雄, 大和淳司, 森島繁生, 陳キュウ. “GANsにおける小規模生成器実現のための敵対的蒸留”, WebDB Forum 2019, P-43, Tokyo, Sept 2019. (Poster, 最優秀学生ポスター発表賞)
星 泰成, 綱島 秀樹, 陳 キュウ. “蒸留を用いたGANの計算量削減手法”, WebDB Forum 2019, P-44, Tokyo, Sept 2019.
吉岡明信, 綱島秀樹, 陳キュウ. “料理画像を用いた味の推定手法”, 電子情報通信学会総合大会, D-12-42, Tokyo, Mar 2019.
綱島秀樹, 佐藤祥, 星泰成, 陳キュウ. “ディープラーニングによる暗所での顔認証手法の検討”, 電子情報通信学会総合大会, D-12-56, Tokyo, Mar 2018.
星泰成, 佐藤祥, 綱島秀樹, 陳キュウ. “敵対的生成ネットワークを用いた顔画像生成手法”, 電子情報通信学会総合大会, D-12-62, Tokyo, Mar 2018.
Award
IPSJ2023 学生奨励賞 (Student Scientist Award) Mar 2023. (視覚情報に基づくタスク指向型対話における人間の応答に対する間違い指摘の検討)
IPSJ2023 学生奨励賞 (Student Scientist Award) Mar 2023. (複数解像度で画像を生成可能な拡散確率モデル)
The 25th Meeting on Image Recognition and Understanding (MIRU2022), インタラクティブセッション賞 (Outstanding Poster Presentation Award), July 2022.
The 25th Meeting on Image Recognition and Understanding (MIRU2022), MIRU論文評価貢献賞 (Outstanding Reviewer Award), July 2022.
The 24th Meeting on Image Recognition and Understanding (MIRU2021), 学生優秀賞 (Student Best Paper Award), July 2021.
2020年度工学院大学大学表彰 (University Award in Kogakuin University 2020).
18th International Symposium on Advanced Technology (ISAT-18), Excellent Oral Paper Presentation Award, Dec 2019.
WebDB Forum 2019, 最優秀学生ポスター発表賞 (Student Best Poster Presentation Award), Sept 2019.
綱島秀樹, 中間康文, 枇々木裕太, “2019年度 Fintech Data Championship”, 特別賞 (Special Award), Mar 2019.
Review
The 26th Meeting on Image Recognition and Understanding (MIRU2023)
The 25th Meeting on Image Recognition and Understanding (MIRU2022), MIRU論文評価貢献賞 (Outstanding Reviewer Award)
Thirty-ninth International Conference on Machine Learning (ICML2022)
Talks
Paper introduction of MR-VAE
【ICLR2023論文解説】Disentanglementは自由自在!?MR-VAEの紹介【cvpaper.challengeコラボ企画】
Paper introduction of StyleNeRF
Talk of the Histry of GANs(【スキルアップAIキャンプ】第79回『 生成モデルはまだまだ進化している! GAN の研究動向紹介』)
Paper introduction of StyleNeRF
Panel Discussion on cvpaper.challenge conference 2021 (CCC2021)
Paper introduction of CVPR2020
I introduced GANs papers of the CVPR2020 as the student representative of AI-SCHOLAR on July 2020.
日本ディープラーニング協会主催 CVPR 2020 技術報告会Paper introduction of ICLR2020
I introduced a object-oriented representation learning paper of the ICLR2020 on June 2020.
ICLR2020オンライン読み会
Presentation MaterialSurvey and presentation of Disentanglement
I surveyed the disentanglement research domain and gave a presentation on Oct 2019.
NLP/CV SoTA Survey Challenge #4
Presentation MaterialIntroduction of GPU parallelization
I introduced the GPU parallelization of Chainer in AIST on Sept 2019.
Presentation MaterialPaper introduction of CVPR2019
I introduced a GANs paper of the CVPR2019 in AIST on Aug 2019.
Presentation Material
Teaching
Generative Adversarial Networks Class
I supervised the GANs class and am teaching students in the GANs class.
Skillup AI GANs class
Service, leadership, others
cvpaper.challenge Head Quarter
Generations Group Leadercvpaper.challenge aims we make novel research trends by reflecting the current computer vision research domain, such as submitting papers to top tier conferences, surveying papers of computer vision conferences, and so on.
I attend cvpaper.challenge as an executive.
Moreover, I lead the Generations group in cvpaper.challenge, which deals generative models.
cvpaper.challengeComprehensive Survey in cvpaper.challenge
We conducted the comprehensive surveys of top tier computer vision conferences, such as CVPR, ICCV, and ECCV every year.
Comprehensive Survey MaterialMeta Survey in Generations of cvpaper.challenge
We conducted the surveys as for generative models and I led this project.
Disentanglement, I2I, Image Manipulation, Latent Space of GANs
Unpaired I2I, NeRF, Object-oriented Representation Learning, Domain Adaptation