I'm a researcher at Preferred Networks, Inc. Kaggle Master (profile). I love Chainer, computer vision, music, films, and wine 🍷

News

• 2021 Jan. Our paper, "Joint Search of Data Augmentation Policies and Network Architectures", has been accepted for AAAI Workshop 2021 "W17: Learning Network Architecture during Training"!
• 2020 Sep. Our paper, "Addressing Class Imbalance in Scene Graph Parsing by Learning to Contrast and Score", has been accepted for ACCV 2020!
• 2020 May. Our paper, "Train Sparsely, Generate Densly: Memory-Efficient Unsupervised Training of High-Resolution Temporal GAN", has been accepted for International Journal of Computer Vision (IJCV)!
• 2019 June. Selected for "CVPR 2019 Outstanding Reviewers"!

Research Experience

• 2016.9 -   Researcher at Preferred Networks, Inc.
  Research for computer vision applications, deep learning framework (Chainer) development
• 2016.2 - 2016.8   Software Engineer (Contractor) at Facebook, Inc.
  Conducted research on large-scale machine learning applications
• 2015.4 - 2016.2   Researcher at Keio Leading-edge Laboratory
  Conducted research on vision-based object detection for personal robots
• 2013.6 - 2014.5   Visiting student researcher at University of California, Berkeley
  Conducted research on environment recognition for guide robots for visually impaired people

Education

• 2016.3   Ph.D in Engineering at Graduate School of Science and Technology, Keio University
  My dissertation is about road and building extraction from aerial imagery using convolutional neural networks. All codes for performing sementic segmentation to aerial imagery is available here: Caffe version and Chainer version.
• 2012.4 - 2015.3   Ph.D student at Graduate School of Science and Technology, Keio University
• 2012.3   Master of Science in Engineering at Graduate School of Science and Technology, Keio University
• 2010.3   Bachelor of Engineering at Department of Electronics and Electrical Engineering, Faculty of Science and Technology, Keio University

Society Activity

• Reviewer: Since 2019~, CVPR, ICCV, ECCV, AAAI, ICLR, etc.
• Program Committee of Workshop on Positive and Negative Adversaries in Multimedia, ACM Multimedia 2018

Research Interest

• Improving the performance of convolutional neural networks on computer vision tasks, e.g., image recognition, object detection, and semantic segmentation
• Generative Adversarial Nets (GAN) for time-series data generation

1. Masaki Saito, Shunta Saito, Masanori Koyama, Sosuke Kobayashi, "Train Sparsely, Generate Densely: Memory-Efficient Unsupervised Training of High-Resolution Temporal GAN", International Journal of Computer Vision (IJCV), DOI, 2020
2. Shunta Saito, Ryota Arai, Yoshimitsu Aoki, "Seamline Determination Based on Semantic Segmentation for Aerial Image Mosaicking", IEEE Access, Vol. 3, pp. 2847-2856, 2015
3. Shunta Saito, Takayoshi Yamashita, Yoshimitsu Aoki, "Multiple Object Extraction from Aerial Imagery with Convolutional Neural Networks", Journal of Imaging Science and Technology, Vol. 60, No. 1, pp. 10402-1-10402-9, 2015

1. Taiga Kashima, Yoshihiro Yamada, Shunta Saito, "Joint Search of Data Augmentation Policies and Network Architectures", AAAI workshop 2021 (W17: Learning Network Architecture during Training), 2021, [arvix]
2. He Huang, Shunta Saito, Yuta Kikuchi, Eiichi Matsumoto, Wei Tang, Philip S. Yu, "Addressing Class Imbalance in Scene Graph Parsing by Learning to Contrast and Score", Asian Conference on Computer Vision (ACCV), 2020, [arxiv]
3. Seiya Tokui, Ryosuke Okuta, Takuya Akiba, Yusuke Niitani, Toru Ogawa, Shunta Saito, Shuji Suzuki, Kota Uenishi, Brian Vogel and Hiroyuki Yamazaki Vincent, "Chainer: a Deep Learning Framework for Accelerating the Research Cycle", KDD Applied Data Science Track, 2019, [arxiv]
4. Azumi Maekawa, Ayaka Kume, Hironori Yoshida, Jun Hatori, Jason Naradowsky, Shunta Saito, "Improvised Robotic Design with Found Objects", Advances in Neural Information Processing Systems (NIPS) Workshop on Machine Learning for Creativity and Design, 2018, [pdf]
5. Shunta Saito*, Tommi Kerola*, Satoshi Tsutsui*, David J. Crandall, "Minimizing Supervision for Free-space Segmentation", IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshop (WAD2018), 2018, [arxiv] (*: These three authors contributed equally.)
6. Satoshi Tsutsui, Tommi Kerola, Shunta Saito, "Distantly Supervised Road Segmentation", International Conference on Computer Vision (ICCV) Workshop (CVRSUAD), 2017, [arxiv]
7. Yusuke Niitani, Toru Ogawa, Shunta Saito, Masaki Saito, "ChainerCV: a Library for Deep Learning in Computer Vision", ACM Multimedia (ACMMM), Open Source Software Competition, 2017, [arxiv]
8. Masaki Saito*, Eiichiro Matsumoto*, Shunta Saito, "Temporal Generative Adversarial Nets with Singular Value Clipping", International Conference on Computer Vision (ICCV), 2017, *equally contributed, [arxiv]
9. Shunta Saito, Takayoshi Yamashita, Yoshimitsu Aoki, “Multiple Objects Extraction from Aerial Imagery with Convolutional Neural Networks,” IS&T International Symposium on Electronic Imaging in the Intelligent Robots and Computer Vision: Algorithms and Techniques Conference (ROBVIS), 17 Feb, 2016
10. Yu Sugawara, Takeshi Morita, Shunta Saito and Takahira Yamaguchi, "An Intelligent Application Development Platform for Service Robots", IROS 2015 Workshop on Multimodel Semantics for Robotic Systems (MuSRobS), 28 September, 2015
11. Siti Nor Khuzaimah Amit, Shunta Saito, Shiori Sasaki, Yoshimitsu Aoki, "Semantic Segmentation and Difference Extraction Via Time Series Aerial Video Camera and Its Application", 36th International Symposium on Remote Sensing (ISRSE), 13 May, 2015
12. Shunta Saito, Yoshimitsu Aoki, "Building and road detection from large aerial imagery", IS&T/SPIE Electronic Imaging, San Francisco (USA), 11 February, 2015
13. Shunta Saito, Makiko Kouchi, Masaaki Mochimaru, Yoshimitsu Aoki, "Model-based 3D human shape estimation from silhouettes for virtual fitting", IS&T/SPIE Electronic Imaging, San Francisco (USA), 5 February, 2014
14. Shunta Saito, Shuntaro Yamazaki, Makiko Kouchi, Masaaki Mochimaru, Yoshimitsu Aoki, "Building 3D Deformable Body Model and Torso Shape Estimation System", 2nd International Digital Human Modeling Symposium, Ann Arbor (USA), 11-13 June, 2013
15. Shuntaro Yamazaki, Yuji Yoshida, Shunta Saito, Makiko Kouchi, Masaaki Mochimaru, Yoshimitsu Aoki, "Homologous Shape Modeling of Human Foot in Motion", Footwear Biomechanics Symposium, Natal (Brasil), 31 July-2 August, 2013
16. Yuji Yoshida, Shunta Saito, Makiko Kouchi, Masaaki Mochimaru, Yoshimitsu Aoki, "Shape completion and modeling of 3D foot shape while walking", International Symposium on Optomechatronic Technologies, pp. 1-3, Paris (France), 29-31 October, 2012
17. Shunta Saito, Makiko Kouchi, Masaaki Mochimaru, Yoshimitsu Aoki, "Simple system for 3D body shape estimation", IEEE 1st Global Conference on Consumer Electronics (GCCE), pp. 203-205, Tokyo (Japan), 2-5 October, 2012
18. Shunta Saito, Makiko Kouchi, Masaaki Mochimaru, Yoshimitsu Aoki, "Body trunk shape estimation from silhouettes by using homologous human body model", 2nd International Conference on 3D Body Scanning Technologies, pp. 130-137, Lugano (Switzerland), 25-26 October, 2011
19. Yuji Yoshida, Shunta Saito, Makiko Kouchi, Masaaki Mochimaru, Yoshimitsu Aoki, "Shape completion and modeling of 3D foot shape while walking using homologous model fitting", 2nd International Conference on 3D Body Scanning Technologies , pp. 270-276, Lugano (Switzerland), 25-26 October, 2011
20. Shunta Saito, Makiko Kouchi, Masaaki Mochimaru, Yoshimitsu Aoki, "Body Shape Estimation by Homologous Human Body Model Fitting", 8th France-Japan Congress on Mecatronics , pp.361-364, Yokohama (Japan), 22-24 November, 2010
21. Shunta Saito, Makiko Kouchi, Masaaki Mochimaru, Yoshimitsu Aoki, "Image Measurement of Hand Dimensions and Model-Based Human Hand Posture Estimation", 36th Annual Conference of the IEEE Industrial Electronics Society , pp.1127-1131, Phoenix (USA), 7-10 November, 2010

1. 齋藤 俊太, 山下 隆義, 青木 義満, "Convolutional Neural Networkを用いた航空画像からの道路・建物抽出", IS3-20, 画像センシングシンポジウム, 神奈川, 6月10日, 2016
2. 工藤 康統, 齋藤 俊太, 青木 義満, "Convolutional Neural Networksを用いた人物周辺の物体とシーンを考慮した行動認識", IS2-15, 画像センシングシンポジウム, 神奈川, 6月9日, 2016
3. 山中 正雄, 齋藤 俊太, 青木 義満, "人体の部位の位置の大域的構造に基づく車両内乗員の行動認識手法", IS1-18, 画像センシングシンポジウム, 神奈川, 6月8日, 2016
4. 齋藤 俊太, 山下 隆義, 青木義満, "Convolutional Neural Networkを用いた航空画像からの道路・建物抽出", 画像センシングシンポジウム, 神奈川, 6月12日, 2015
5. 齋藤 俊太, 青木 義満, "自律分散型画像センシングモジュール群からなるロボットの知覚システムの検討", 第29回人工知能学会全国大会, 函館, 5月30日, 2015
6. 山口 高平, 中野 有紀子, 斎藤 英雄, 森田 武史, 青木 義満, 萩原 将文, 斎藤 俊太, "知能共進化のための実践知能アプリケーションプラットフォームPRINTEPS", 第29回人工知能学会全国大会, 函館, 5月30日, 2015
7. 横田 宏行, 杉山 史典, 齋藤 俊太, 青木 義満, "空中写真からの家屋抽出に関する研究", 日本写真測量学会 秋季学術講演会, 11月20日, 2014
8. 北條 さやか, 齋藤 俊太, 青木 義満, "ファッションにおける曖昧な趣味趣向の理解・表現", 電気学会電子・情報・システム部門大会, 9月4日, 2013
9. 齋藤 俊太, 青木 義満, "視覚認知システムにおける知覚と推論 Perception and Inference in Visual Cognition Systems", 計測自動制御学会 第87回パターン計測部会研究会, 8月2日, 2012, PDF, SLIDE
10. 齋藤 俊太, 河内 まき子, 持丸 正明, 青木 義満, "姿勢変動を考慮したモデルベース人体3次元形状推定", ViEW2011 ビジョン技術の実利用ワークショップ, 神奈川, 12月8日, 2011
11. 齋藤 俊太, 河内 まき子, 持丸 正明, 青木 義満, "モデルベース人体体幹部3次元形状推定",コンピュータビジョンとイメージメディア研究発表会, 函館, 9月5日, 2011
12. 齋藤 俊太, 持丸 正明, 河内 まき子, 青木 義満, "シルエット画像を用いたモデルベース人体 体幹部形状推定", 画像センシングシンポジウム, 神奈川, 6月9日, 2011
13. 竹内 冠太, 相島 雅樹, 齋藤 俊太, 鈴木 晋, 植木 淳朗, 杉浦 一徳, "プロジェクトにおける創発的協調作業モデルの提案", 情報処理学会研究報告 情報システムと社会環境研究報告, 2011-IS-116(5), pp.1-8, 2011-05-30
14. 齋藤 俊太, 河内 まき子, 持丸 正明, 青木 義満, "シルエット画像を用いたモデルベース人体 体幹部三次元形状推定", ViEW2010 ビジョン技術の実利用ワークショップ, 神奈川, 12月9日, 2010
15. 齋藤 俊太, 高橋 拓也, 持丸 正明, 河内 まき子, 青木 義満, "手部寸法の画像計測によるデジタルハンドの生成と把持姿勢推定", 画像センシングシンポジウム, 神奈川, 6月10-12日, 2009
16. 齋藤 俊太, 持丸 正明, 河内 まき子, 青木 義満, "手部寸法の画像計測と把持姿勢推定への応用", 精密工学会, 埼玉, 3月16-18日, 2010
17. 齋藤 俊太, 高橋 拓也, 持丸 正明, 河内 まき子, 青木 義満, "製品設計支援のための手部寸法自動計測と把持姿勢推定", 画像センシングシンポジウム, 神奈川, 6月9-11日, 2010

1. SSIIオーディエンス賞: 工藤 康統，齋藤 俊太，青木 義満, "Convolutional Neural Networksを用いた人物周辺の物体とシーンを考慮した行動認識", IS2-15, 画像センシングシンポジウム, 神奈川, 6月9日, 2016

1. 齋藤 俊太, "敵対的生成ネットワークによる画像生成", 光学, 第 47 巻 第 12 号 (2018年12月)
2. [翻訳] 齋藤 俊太, "『物質と記憶』と深層学習", ベルクソン『物質と記憶』を診断する, [amazon]
3. 中村 薫, 齋藤 俊太, 宮城 英人, "KINECT for Windows SDKプログラミング C++編", 秀和システム, 2012年6月19日刊行, ISBN-10: 4798033707
4. 齋藤俊太, 青木義満, "シルエット画像を用いたモデルベース人体体幹部形状推定", 画像ラボ, 日本工業出版, V.23/No.3/P.19-24, 2012/03/10