Kevin Judd
Kevin Judd

Kevin joined ORI in 2016 as a DPhil student at Somerville College after having spent a year in Oxford as part of the CDT in Autonomous Intelligent Machines and Systems program.

He is particularly interested in robotic vision systems, and is currently working on simultaneous ego- and third-party motion estimation and tracking using stereo vision.

In his free time, he enjoys running and playing football.


  • Download associated data
  • Watch associated video
  • Find paper by Direct Object Identifier
  • Find paper on arXiv (open access)
  • Find updated paper on arXiv (open access)
  • Find paper on Google Scholar
  • Journal Papers

    1. K. M. Judd, J. D. Gammell. “The Oxford Multimotion Dataset: Multiple SE(3) motions with ground truth.” IEEE Robotics and Automation Letters (RA-L), 4(2):800–807, Apr. 2019. Presented at ICRA 2019.

    Conference Papers


    1. K. M. Judd, J. D. Gammell, P. Newman. “Multimotion visual odometry (MVO): Simultaneous estimation of camera and third-party motions.” in Proceedings of the IEEE/RSJ international conference on intelligent robots and systems (IROS), pp. 3949–3956, Madrid, Spain, 1–5 Oct. 2018.


    1. K. M. Judd, J. D. Gammell. “SE(3) multimotion estimation through occlusion.” in Proceedings of the long-term human motion prediction (LHMP) workshop, IEEE international conference on robotics and automation (ICRA), MontrĂ©al, QC, Canada, 24 May 2019.
    2. K. M. Judd, J. D. Gammell, P. Newman. “Visual multimotion estimation.” in Joint industry and robotics CDTs symposium (JIRCS), Bristol, UK, 4 Jun. 2018.

    Technical Reports

    1. K. M. Judd, J. D. Gammell. “Occlusion-robust MVO: Multimotion estimation through occlusion via motion closure.” Technical Report, Estimation, Search, and Planning (ESP) Research Group, University of Oxford, May 2019.