Tatsunori Taniai
谷合 竜典

Email: tatsunori.taniai [at] sinicx.com
CV   DBLP   GoogleScholar   ResearchGate   SemanticScholar   GitHub

I am a senior researcher at OMRON SINIC X Corporation in Tokyo. I am interested in geometric and photometric 3D vision and physics-informed machine learning.

Call for interns: I am looking for an intern who can intensively work with me on a computer vision research problem using deep learning, aiming at publishing a paper at CVPR/ICCV/ECCV/ICML/NIPS.


Affilication

2019 - Senior Researcher, OMRON SINIC X Corporation

Knowledge Computing Group

Research
projects
 

Crystalformer: Infinitely Connected Attention for Periodic Structure Encoding

Tatsunori Taniai, Ryo Igarashi, Yuta Suzuki, Naoya Chiba, Kotaro Saito, Yoshitaka Ushiku, Kanta Ono
ICLR 2024
[Preprint] [Project] [Code]

 

Risk-aware Path Planning via Probabilistic Fusion of Traversability Prediction for Planetary Rovers on Heterogeneous Terrains

Masafumi Endo, Tatsunori Taniai, Ryo Yonetani, Genya Ishigami
ICRA 2023
[Preprint] [Project] [Code]

 

Self-supervised learning of materials concepts from crystal structures via deep neural networks

Yuta Suzuki, Tatsunori Taniai, Kotaro Saito, Yoshitaka Ushiku, Kanta Ono
Machine Learning: Science and Technology 2022
[Open Access] [Code]

 

Quasistatic contact-rich manipulation via linear complementarity quadratic programming

Sotaro Katayama, Tatsunori Taniai, Kazutoshi Tanaka
IROS 2022
[Preprint] [Project] [Code]

 

Path Planning using Neural A* Search

Ryo Yonetani*, Tatsunori Taniai*, Mohammadamin Barekatain, Mai Nishimura, Asako Kanezaki
ICML 2021
[Preprint] [Project] [Code]

 

Neural Inverse Rendering for General Reflectance Photometric Stereo

Tatsunori Taniai, Takanori Maehara.
ICML 2018
[Preprint] [Project] [Code]

 

Continuous 3D Label Stereo Matching using Local Expansion Moves

Tatsunori Taniai, Yasuyuki Matsushita, Yoichi Sato, Takeshi Naemura.
TPAMI 2018
[Preprint] [Project] [Code]

 

Semi-Global Stereo Matching with Surface Orientation Priors

Daniel Scharstein, Tatsunori Taniai, Sudipta N. Sinha.
3DV 2017 Spotlight
[Preprint (extended)] [Poster] [Slides]

 

Fast Multi-frame Stereo Scene Flow with Motion Segmentation

Tatsunori Taniai, Sudipta N. Sinha, Yoichi Sato.
CVPR 2017
[Preprint] [Preprint (extended)] [Poster] [Project] [Demo]

 

Joint Recovery of Dense Correspondence and Cosegmentation in Two Images

Tatsunori Taniai, Sudipta N. Sinha, Yoichi Sato.
CVPR 2016
[Preprint] [Preprint (extended)] [Poster] [Project] [Demo]

 

Superdifferential Cuts for Binary Energies

Tatsunori Taniai, Yasuyuki Matsushita, Takeshi Naemura.
CVPR 2015
[Preprint] [ExtAbst] [Poster] [Project] [Code]

 

Graph Cut based Continuous Stereo Matching using Locally Shared Labels

Tatsunori Taniai, Yasuyuki Matsushita, Takeshi Naemura.
CVPR 2014
[Preprint] [Poster] [Project]

 

Image Segmentation using Dual Distribution Matching

Tatsunori Taniai, Viet-Quoc Pham, Keita Takahashi, Takeshi Naemura.
BMVC 2012 Oral
[Preprint] [Supp] [ExtAbst] [Slides] [VideoLecuture] [Project]



Research interests

3D Reconstruction

  • Accurate stereo vision that estimates dense per-pixel 3D planes (e.g. CVPR 2014)
  • A fusion approach of multiview stereo and photometric stereo.

MRF Optimization

  • Apply graph cuts to optimization of MRFs with higher-order clieques for image segmentation (e.g. CVPR 2015, BMVC 2012)
  • Efficient inference of MRFs with a very large or continuous label space (e.g. CVPR 2014)

Deep Learning

  • Interesting application of deep learning for computer vision problems (e.g. ICML 2018)
Education

Experience

Working History

  • Senior Researcher (Apr. 2019 - present, Perception Group, OMURON SINIC X Corp., Japan)
  • Visiting Researcher (Jul. 2019 - Nov. 2020, Discrete Optimization Unit led by Dr. Takanori Maehara, RIKEN AIP Center, Japan)
  • Special Postdoctoral Researcher / 特別研究員 (May 2017 - Mar. 2019, Discrete Optimization Unit led by Dr. Takanori Maehara, RIKEN AIP Center, Japan)
  • Research Assistant Part-timer / 研究支援パートタイマー (Apr. 2017, Discrete Optimization Unit led by Dr. Takanori Maehara, RIKEN AIP Center, Japan)
  • JSPS Young Research Fellow / 日本学術振興会 特別研究員 (DC1) (Apr. 2014 - Mar. 2017 at the University of Tokyo, Japan)

Research Internship and Visiting

  • Microsoft Research (May - Aug. 2016 at Redmond, USA)
    Twelve weeks of research internship supevised by Dr. Sudipta Sinha.
    Part of the internship achievements has been published as a CVPR 2017 paper.

  • Microsoft Research Asia (Jan.- Apr. 2016 at Beijing, China)
    Three months of visiting research supevised by Dr. David Wipf.

  • Microsoft Research (June - Sep. 2015 at Redmond, USA)
    Twelve weeks of research internship supevised by Dr. Sudipta Sinha.
    Part of the internship achievements has been published as a CVPR 2016 paper.

  • Microsoft Research Asia (Dec. 2012 - Apr. 2013 at Beijing, China)
    Four months of research internship supevised by Dr. Yasuyuki Matsushita.
    Part of the internship achievements has been published as a CVPR 2014 paper.

Service to Community

  • Conference Reviewer: ECCV '20, CVPR '23, '22, '20, '18, ICCV '23, '19, '17, 3DV '18, '14
  • Journal Reviewer: TPAMI '23, '19, '18, IJCV '18, CVIU '17, IMAVIS '16, IEEE TIP '18, '15.
Awards/
honours
  • Outstanding Reviewer Awards
    from CVPR 2023, ECCV 2020, CVPR 2020, CVPR 2018.

  • IPSJ Yamashita SIG Research Award / 山下記念研究賞
    from Information Processing Society of Japan (2018) [Link]

  • Dean's Award for Best Doctoral Thesis / 東京大学 情報理工学系研究科長賞(博士)
    from Graduate School of Information Science and Technology, the University of Tokyo (2017)

  • Microsoft Research Asia Fellowship 2015 (10,000 USD)
    from Microsoft Research Asia for 13 PhD studens studying CS in Asia. (2015-2017).

  • JASSO Scholarship for Top 10% Excellent Master Students (2112,000 JPY)
    / 日本学生支援機構 奨学金 全額返済免除 (211.2万円)
    from Japan Student Services Organization (2014)

  • Dean's Award for Best Master Thesis / 東京大学 情報理工学系研究科長賞(修士)
    from Graduate School of Information Science and Technology, the University of Tokyo (2014)

  • Dean's Award for Best Bachelor Thesis / 東京大学 工学部長賞
    from Faculty of Engineering, the University of Tokyo (2012)

Skills
  • 15+ years of programming experiences in C++, C#, Java
  • Visual computing using OpenCV and MATLAB
  • GPGPU computing using OpenCL and CUDA
  • Academic literacy & conversation skills in English (TOEIC 930 in May 2011)


Pubilcations

Books and Chapters

  • Tatsunori Taniai. Binocular Stereo. In Ikeuchi K. (eds) Computer Vision. Springer, Cham, 2020.
    [Preprint]

Journal Papers

  • Naoya Chiba, Yuta Suzuki, Tatsunori Taniai, Ryo Igarashi, Kotaro Saito, Yoshitaka Ushiku, Kanta Ono. Neural structure fields with application to crystal structure autoencoders. Communication Materials, 2023 (accepted)
    [Open Access] [Code]

  • Yuta Suzuki, Tatsunori Taniai, Kotaro Saito, Yoshitaka Ushiku, Kanta Ono. Self-supervised learning of materials concepts from crystal structures via deep neural networks. Machine Learning: Science and Technology, 2022. (accepted)
    [Open Access] [Code]

  • Tatsunori Taniai, Yasuyuki Matsushita, Yoichi Sato, Takeshi Naemura. Continuous 3D Label Stereo Matching using Local Expansion Moves. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI 2018), vol. 40, no. 11, pp. 2725--2739, 2018.
    [Preprint] [Project] [Code]

  • 谷合竜典, ファン・ヴェト・クォク, 高橋桂太, 苗村健. 前景・背景色分布の同時マッチングによる画像セグメンテーション. 電子情報通信学会和文論文誌D, 画像の認識・理解特集号, vol. J96-D, no. 8, pp. 1764--1777, 2013. (an extended version of our BMVC 2012's paper)

International Conference Papers

  • Tatsunori Taniai, Ryo Igarashi, Yuta Suzuki, Naoya Chiba, Kotaro Saito, Yoshitaka Ushiku, Kanta Ono
    Crystalformer: Infinitely Connected Attention for Periodic Structure Encoding. In Proceedings of The Twelfth International Conference on Learning Representations (ICLR 2024), Vienna, Austria, 2024.
    [Preprint] [Project] [Code]

  • Masafumi Endo, Tatsunori Taniai, Ryo Yonetani, Genya Ishigami. Risk-aware Path Planning via Probabilistic Fusion of Traversability Prediction for Planetary Rovers on Heterogeneous Terrains. In Proceedings of the 2023 IEEE International Conference on Robotics and Automation (IROS 2023), pp. 11852--11858, London, UK, 2023.
    [Preprint] [Project] [Code]

  • Sotaro Katayama, Tatsunori Taniai, Kazutoshi Tanaka. Quasistatic contact-rich manipulation via linear complementarity quadratic programming. In Proceedings of the 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2022), pp. 203–-210, Kyoto, Japan, 2022.
    [Preprint] [Project] [Code]

  • Ryo Yonetani*, Tatsunori Taniai*, Mohammadamin Barekatain, Mai Nishimura, Asako Kanezaki. Path Planning using Neural A* Search. In Proceedings of the International Conference on Machine Learning (ICML 2021), a virtual conference, 2021. (acceptance rate 21%)
    [Preprint] [Project] [Code]

  • Tatsunori Taniai, Takanori Maehara. Neural Inverse Rendering for General Reflectance Photometric Stereo. In Proceedings of the International Conference on Machine Learning (ICML 2018), pp. 4864--4873, Stockholm, Sweden, 2018. (acceptance rate 25%)
    [Preprint] [Project] [Code]

  • Daniel Scharstein, Tatsunori Taniai, Sudipta N. Sinha. Semi-Global Stereo Matching with Surface Orientation Priors. In Proceedings of the 5th International Conference on 3D Vision (3DV 2017 Spotlight), pp. 215--224, Qingdao, China, 2017.
    [Preprint (extended)] [Poster] [Slides]

  • Tatsunori Taniai, Sudipta N. Sinha, Yoichi Sato. Fast Multi-frame Stereo Scene Flow with Motion Segmentation. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2017), pp. 6891--6900, Honolulu, Hawaii, USA, 2017. (acceptance rate 30%)
    [Preprint] [Preprint (extended)] [Poster] [Project]

  • Tatsunori Taniai, Sudipta N. Sinha, Yoichi Sato. Joint Recovery of Dense Correspondence and Cosegmentation in Two Images. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2016), pp. 4246--4255, Las Vegas, NV, USA, 2016. (acceptance rate 30%)
    [Preprint] [Preprint (extended)] [Poster] [Project]

  • Tatsunori Taniai, Yasuyuki Matsushita, Takeshi Naemura. Superdifferential Cuts for Binary Energies. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2015), pp. 2030--2038, Boston, MA, USA, 2015. (acceptance rate 28%)
    [Preprint] [ExtAbst] [Project]

  • Tatsunori Taniai, Yasuyuki Matsushita, Takeshi Naemura. Graph Cut based Continuous Stereo Matching using Locally Shared Labels. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2014), pp. 1613--1620, Columbus, OH, USA, 2014. (acceptance rate 30%)
    [Preprint] [Poster] [Project] [Code]

  • Tatsunori Taniai, Viet-Quoc Pham, Keita Takahashi, Takeshi Naemura. Image Segmentation using Dual Distribution Matching. In Proceedings of the 23rd British Machine Vision Conference (BMVC 2012), pp. 74.1--74.11, Surrey, UK, 2012. (oral presentation; acceptance rate 8%)
    [Preprint] [Supp] [ExtAbst] [Slides] [VideoLecuture] [Project]

Invited Talks

  • Tatsunori Taniai, Sudipta N. Sinha, Yoichi Sato. Fast Multi-frame Stereo Scene Flow with Motion Segmentation (CVPR 2017). The 20th Meeting on Image Recognition and Understanding (MIRU 2017), Hiroshima, Japan, August 10th, 2017.

  • Tatsunori Taniai. Joint Recovery of Dense Correspondence and Cosegmentation in Two Images. The Workshop on Vision, Learning, and Cognition in Microsoft Research Asia Ph.D. Forum 2016, Microsoft office, Beijing, China, September 20th, 2016.

  • Tatsunori Taniai, Sudipta N. Sinha, Yoichi Sato. Joint Recovery of Dense Correspondence and Cosegmentation in Two Images (CVPR 2016). The 19th Meeting on Image Recognition and Understanding (MIRU 2016), IS2-15, Shizuoka, Japan, August 4th, 2016.

  • Tatsunori Taniai. Solving Segmentation and Dense Correspondence Problems using Graph Cuts (画像領域・対応点推定問題へのグラフカットの適用). The 1st CREST Symposium on Random Fields and Deep Learning , Waseda Univ., Tokyo, Japan, January 13th, 2016. (Organizers: Prof. Hiroshi Ishikawa & Prof. Takayuki Okatani)
    [Slides]

  • Tatsunori Taniai, Yasuyuki Matsushita, Takeshi Naemura. Superdifferential Cuts for Binary Energies (CVPR 2015). The 18th Meeting on Image Recognition and Understanding (MIRU 2015), IS1-10, Osaka, Japan, July 28th, 2015.

  • Tatsunori Taniai, Yasuyuki Matsushita, Takeshi Naemura. Graph Cut based Continuous Stereo Matching using Locally Shared Labels (CVPR 2014). The 17th Meeting on Image Recognition and Understanding (MIRU 2014), IT1-1, Okayama, Japan, July 29th, 2014.

Domestic Conference Papers

  • 谷合竜典, 苗村健. 画素位置を埋め込んだ双色分布マッチングによる画像セグメンテーション. 画像の認識・理解シンポジウム (MIRU2013)予稿集, SS3-31, 東京, 2013.

  • 谷合竜典, ファン・ヴェト・クォク, 高橋桂太, 苗村健. 前景および背景色分布の同時マッチングによる画像セグメンテーション. 画像の認識・理解シンポジウム (MIRU2012)予稿集, OS14-02, 福岡, 2012.

Thesis

  • Tatsunori Taniai. Discrete Inference Approaches to Image Segmentation and Dense Correspondence. Doctoral Thesis. March 2017. The University of Tokyo. Adviser: Yoichi Sato.
    [PDF] [UT-Repo]

  • Tatsunori Taniai. Applying Graph Cuts to MAP Estimation of Continuous and Higher-Order Markov Random Fields. Master's Thesis. March 2014. The University of Tokyo. Adviser: Takeshi Naemura.
    [PDF]

  • Tatsunori Taniai. 前景・背景の大域的な色分布の同時マッチングによる画像セグメンテーション
    (Foreground Background Image Segmentation using Dual Distribution Matching). Bachelor's Thesis. March 2012. The University of Tokyo. Adviser: Takeshi Naemura.
    [PDF] [Note]

Links