Yijun Yuan 元祎君

Yijun is interested in Robotics and Machine learning. He has basic and sufficient knowledge on Computer vision, Machine learning and Robotics. His projects cover topics related to Image regression, Topological representation, Kernel learning and Robot learning. Recently he mainly works on Robot learning for arms.

Education

  • Sept. 2018 – Present   M.E., Computer Science and Technology, School of Information Science and Technology, ShanghaiTech University, China.
  • Sept. 2014 – Jun. 2018   B.E., Computer Science and Technology, School of Information Science and Technology, ShanghaiTech University, China.

Experience

Teaching

  • Spring 2018  Teaching Assistant, Computer Architecture I.

Research

  • Fall 2016 - Aug. 2017   Computer Vision (Retina image segmentation, Crowd Counting).
  • Sept. 2017 - Aug. 2018   Mapping, Robotics(Fast Gaussian Process Occupancy Mapping(Accepted to ICARCV2018), Incrementally building topology graphs via distance maps(Submitted to ICRA2019), Topological Area Graph Generation and its Application to Path Planning(Submitted to ICRA2019)).
  • May. 2018 - Oct. 2018   Machine Learning (Deep Kernel Learning with Randomized Sketches(Waiting for submission)).
  • Oct. 2018 - Nov. 2018   Machine Learning (CBCT Calibration).
  • Oct. 2018 - present   Robot Learning (RL/IL on Arms (Motion Planning and Learning Inverse Kinematic) ).

Awards

  • 2016 Dean’s Scholarship ShanghaiTech University.
  • 2017 Excellent Scholarship, ShanghaiTech University.
  • 2018 Fan’s Favorite Prize, NO.4 in total score, Best on HPCG and Tensorflow, ISC high performance competition, Frankfurt, Germany.

Research Publications

  • 1 Jiawei, H., Yuan, Y. & Schwertfeger, S. (2018). Topological area graph generation and its application to path planning. arXiv preprint arXiv:1811.05113.
  • 2 Yuan, Y., Haofei, K. & Schwertfeger, S. (2018). Fast gaussian process occupancy maps. In 15th incremental conference on control, automation, robotics, and vision.
  • 3 Yuan, Y. & Schwertfeger, S. (2018). Incrementally building topology graphs via distance maps. arXiv preprint arXiv:1811.01547.

Linkedin

CV