Param Hanji

Param Hanji

PhD student in Computer Graphics and Vision

Curriculum Vitae

If you prefer a concise résumé, please see find it here.

Education

  • PhD in Computer Vision, University of Cambridge, 2023 (expected). Funded by the ERC Grant Horizon 2020: Project Eyecode.
  • B.Tech. in Information Technology, National Institute of Technology Karnataka, India, 2018.

Work experience

  • Oct 2022 - Present: Research Associate at University of Cambridge
    • Supervisor: Dr Cengiz Öztireli
    • 3D generative modelling with pointclouds and/or implicit neural representations
    • Helped design and deliver L355: Machine Visual Perception
    • Supervising a year-long MPhil. and Part II project in generative modelling with diffusion methods
  • Feb 2019 - Sept 2022: Research Assistant at University of Cambridge
    • Supervisor: Dr Rafał Mantiuk
    • Working to capture and render images on a novel multi-focalplane, stereoscopic, high dynamic range display. We aim to produce images that will successfully pass the Visual Turing Test.
    • Responsible for reconstructing HDR images and estimating per-view depth maps. MLE-based estimators obtained by assuming a statistical camera noise model produced the best HDR images. Preliminary results were presented at an ECCV workshop, with a more elaborate journal version currently under review.
    • Also worked on optical flow, differentiable rendering, neural view synthesis
  • Sept 2020 - Feb 2021: Part-time project with Huawei Research, Munich
    • Exploring the effect of different tone-curves (encoding functions) on the performance of state-of-the-art Computer Vision methods.
    • Conducted an evaluation of state-of-the-art face and object detectors as well as optical flow with the help of a new adversarial illumination dataset. The dataset and preliminary study were accepted to be published in the Journal of Imaging Science and Technology (JIST).
  • July 2018 - Jan 2019: Research Assistant at National University of Singapore
    • Supervisor: Prof. Gary Tan
    • Involved in the development of agent-based traffic simulator, at the Singapore-MIT Alliance for Research and Technology (SMART) Lab. Contributed to the first release of SimMobility.
    • Worked to improve the runtime performance of an on-demand vehicle controller that handles shared taxi requests. On average, the simulation time was reduced to a third of the original.

Publications

Supervisions

Please see Teaching