RIKEN Center for Advanced Intelligence Project Robotics for Infrastructure Management Team

Team Leader: Takayuki Okatani (D.Eng.)

Japanese Page

Research Summary

There is urgent demand in Japan for making it possible to more efficiently perform inspection of infrastructure such as bridges. Aiming at automating these work by robots, we are studying a variety of AI and Robotic technologies that are necessary for this aim. These include vision-based methods for detecting anomalies such as damages/defects of structures; vision-based methods for recognition of unknown three-dimensional environment surrounding robots and their navigation in the environment; basic research of deep learning methods for these tasks; and development of hardware and autonomous flight control of UAVs with special structures that enable them to continue to fly after collision with surfaces in space.

Main Research Fields

Computer Science

Related Research Fields

Engineering
Neuroscience & Behavior
Mathematics

Research Subjects

Robotics for Infrastructure Inspection and Management
Understanding and Improvement of Deep Learning

Selected Publications

Papers with an asterisk(*) are based on research conducted outside of RIKEN.

1.Eisuke Ito and Takayuki Okatani:
“Self-calibration-based Approach to Critical Motion Sequences of Rolling-shutter Structure from Motion”
Proc. Computer Vision and Pattern Recognition (CVPR), (2017)
2.Ken Sakurada, Daiki Tetsuka and Takayuki Okatani:
“Temporal city modeling using street level imagery”
Computer Vision and Image Understanding 157, 55-71 (2017)
3.*Yasushi Akashi and Takayuki Okatani:
“Separation of reflection components by sparse non-negative matrix factorization”
Computer Vision and Image Understanding 146, 77-85 (2016)
4.*Ken Sakurada, Takayuki Okatani, Kris M. Kitani:
“Hybrid macro-micro visual analysis for city-scale state estimation”
Computer Vision and Image Understanding 146, 86-98 (2016)
5.Zhun Sun, Mete Ozay, and Takayuki Okatani:
“Design of Kernels in Convolutional Neural Networks for Image Classification”
Proc. European Conference on Computer Vision (ECCV), 51-66 (2016)
6.Sirion Vittayakorn, Takayuki Umeda, Kazuhiko Murasaki, Kyoko Sudo, Takayuki Okatani, and Kota Yamaguchi:
“Automatic Attribute Discovery with Neural Activations”
Proc. European Conference on Computer Vision (ECCV), 252-268 (2016)
7.*Masaki Saito and Takayuki Okatani:
“Transformation of Markov Random Fields for marginal distribution estimation”
Proc. Computer Vision and Pattern Recognition (CVPR), 797-805 (2015)
8.*Makoto Ozeki and Takayuki Okatani:
“Understanding Convolutional Neural Networks in Terms of Category-Level Attributes”
Proc. Asian Conference on Computer Vision (ACCV) 362-375 (2014)
9.*Masaki Saito, Takayuki Okatani and Koichiro Deguchi:
“Discrete MRF Inference of Marginal Densities for Non-uniformly Discretized Variable Space”
Proc. Computer Vision and Pattern Recognition (CVPR), 57-64 (2013)
10.*Ken Sakurada, Takayuki Okatani, Koichiro Deguchi:
“Detecting Changes in 3D Structure of a Scene from Multi-view Images Captured by a Vehicle-Mounted Camera”
Proc. Computer Vision and Pattern Recognition (CVPR), 137-144 (2013)

Lab Members

Principal investigator

Takayuki Okatani: Team Leader

Core members

Quang Van Nguyen: Postdoctoral Researcher
Zhijie Wang: Postdoctoral Researcher
Hidehiko Sekiya: Visiting Scientist
Masanori Suganuma: Visiting Scientist

Contact Information

Aramaki Aza Aoba, Aoba-ku,
980-8579 Sendai, Miyagi,
Japan
Email: takayuki.okatani [at] riken.jp