Ahmed Abdelkader

Ahmed Abdelkader

Research Scientist

Google Research


My research tackles fundamental challenges in geometry acquisition and modeling, drawing upon deep theoretical insights to devise novel data representations with a track-record of top-tier publications. I thrive in fast-paced multidisciplinary teams, and have a passion for education and mentoring.

I obtained my PhD in Computer Science from the University of Maryland at College Park, where I wrote a thesis on nearest-neighbor searching and was lucky to have David Mount as my advisor. During my PhD, I spent a lot of time at Sandia National Labs working on meshing algorithms with Scott Mitchell and Mohamed Ebeida. Before leaving Maryland, I did a few projects on adversarial robustness in collaboration with Tom Goldstein and some of his bright students. Prior to all that, I studied computer engineering and TA’ed many math/CS undergrad classes at Alexandria University in Egypt.

  • Computational Geometry
  • Computational Photography
  • Machine Learning
  • Ph.D. in Computer Science, 2020

    University of Maryland

  • M.Sc. in Applied Mathematics, 2013

    Alexandria University

  • B.Sc. in Computer Engineering, 2009

    Alexandria University

Sample Publications

(2023). Smooth Distance Approximation. In ESA.

Cite Project LIPIcs PDF Slides

(2020). Detection as Regression: Certified Object Detection by Median Smoothing. In NeurIPS.

Cite Project NeurIPS arXiv Code

(2020). Certified Defenses for Adversarial Patches. In ICLR.

Cite Project OpenReview arXiv Code

(2019). On Realistic Target Coverage by Autonomous Drones. In TOSN.

Cite Project ACM DL arXiv

(2019). Approximate Nearest Neighbor Searching with Non-Euclidean and Weighted Distances. In SODA.

Cite Project SIAM PDF Slides