Dasith de Silva Edirimuni

DasithdeSilvaEdirimuni.jpg

I was formally trained as a physicist with a Master of Science degree from the University of Melbourne (2017) and a Bachelor of Science in Physics with minors in Mathematics and Computer Science (2013) from Hardin-Simmons University, Abilene, TX. I also completed a Master of Information Technology from Swinburne University, Melbourne in 2020. Most recently, I submitted my PhD thesis, focused on the use of neural encoder-decoder models for point cloud filtering/denoising, to Deakin University, Geelong. Shortly thereafter, I was appointed as a Research Fellow at the University of Western Australia in the group of Prof. Ajmal Mian.

My PhD work is on computer vision. More specifically, I work on applications of deep learning to point cloud processing, analysis and generation. I am also interested in the intersection between language and vision.

UPDATE: Currently I am looking to supervise students interested in computer vision problems. Pre-requisites for scholarship applications include outstanding grades in your current Honours or Masters programs. Previous research outputs are also highly valued. Broadly speaking, potential projects will involve applying a variety of modelling approaches for transport problems (such as diffusion, normalizing flows etc.) for 3D generation. Other research avenues include blue-sky research applying 3D representation learning to other physical problems and/or developing new representation learning techniques. Please feel free to reach out to me by email (button at top left corner).

News

Aug 19, 2024 Started my appointment at UWA as a Research Fellow focused on 3D computer vision.
Jul 1, 2024 SemReg: Semantics Constrained Point Cloud Registration paper accepted to ECCV 2024!
Feb 27, 2024 StraightPCF: Straight Point Cloud Filtering paper accepted to CVPR 2024!
Mar 26, 2023 Contrastive Learning for Joint Normal Estimation and Point Cloud Filtering paper accepted to the IEEE TVCG journal!
Feb 27, 2023 IterativePFN: True Iterative Point Cloud Filtering paper accepted to CVPR 2023!
Mar 9, 2022 Deep Point Cloud Normal Estimation Via Triplet Learning paper accepted to ICME, 2022!
Mar 3, 2021 Awarded the University Medal for most outstanding postgraduate student among Swinburne’s 2020 graduating cohort.
Dec 14, 2020 Graduated from Swinburne University with a Master of Information Technology.
Apr 26, 2019 Interaction and structuration of membrane-binding and membrane-excluding colloidal particles in lamellar phases paper accepted to the Soft Matter journal.

Recent publications

  1. straightpcf_before.png
    Dasith de Silva EdirimuniXuequan LuGang Li, Lei Wei, Antonio Robles-Kelly, and Hongdong Li
    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Jun 2024
  2. iterativepfn_before.png
    Dasith de Silva EdirimuniXuequan LuZhiwen ShaoGang LiAntonio Robles-Kelly, and Ying He
    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Jun 2023
  3. joint_learning_before.png
    Dasith de Silva EdirimuniXuequan LuGang Li, and Antonio Robles-Kelly
    IEEE Transactions on Visualization and Computer Graphics, Jun 2024
  4. triplet_normal_before.png
    Weijia Wang, Xuequan LuDasith de Silva EdirimuniXiao Liu, and Antonio Robles-Kelly
    In 2022 IEEE International Conference on Multimedia and Expo (ICME), Jun 2022
  5. lamellar_before.png
    Ruben ZakineDasith de Silva EdirimuniDoru Constantin, Paolo Galatola, and Jean-Baptiste Fournier
    Soft Matter, Jun 2019