Biography

I am a Postdoctoral Fellow at King Abdullah University of Science and Technology (KAUST), Saudi Arabia, specializing in Generative AI. I obtained my Ph.D. degree in Computer Vision and Deep Learning from Linköping University, Sweden under the supervision of Prof. Michael Felsberg and Prof. Fahad Shahbaz Khan.

I have 11 years of diverse interdisciplinary experience through academic and industrial roles across Sweden, Egypt, and Saudi Arabia.

Areas of Expertise

Diffusion Models Vision Foundation Models Large Language Models (LLM) Retrieval-Augmented Generation (RAG) Multi-Agentic Systems Image/Video Object Detection/Segmentation/Tracking Depth Estimation/Completion Uncertainty in Deep Learning Optical Flow Estimation AI for Sports

Updates

Paper Accepted at SIGGRAPH 2025

"PartEdit: Fine-Grained Image Editing using Pre-Trained Diffusion Models" will be presented at SIGGRAPH 2025 in Vancouver, Canada.

March 2025 View Publication

New Preprint Released

"EditCLIP: Representation Learning for Image Editing" was published on Arxiv.

February 2024 View Publication

Paper Accepted at ICLR 2025

"Build-A-Scene: Interactive 3D Layout Control for Diffusion-Based Image Generation" will be presented at ICLR 2025 in Singapore.

Jan 2025 View Publication

Paper Accepted at CVPR 2025

"Zero-Shot Video Semantic Segmentation based on Pre-Trained Diffusion Models" will be presented at CVPR 2025 in Nashville, USA.

Jan 2025 View Publication
Page 1 of 1

Selected Publications

ICLR 2025
Build-A-Scene

Build-A-Scene: Interactive 3D Layout Control for Diffusion-Based Image Generation

Abdelrahman Eldesokey, Peter Wonka

Pre-Print 2025
edit-clip

EditCLIP: Representation Learning for Image Editing

Qian Wang, Aleks Cvejic, Abdelrahman Eldesokey, Peter Wonka

SIGGRAPH 2025
Interactive 3D Layout Control

PartEdit: Fine-Grained Image Editing using Pre-Trained Diffusion Models

Aleks Cvejic*, Abdelrahman Eldesokey*, Peter Wonka

CVPR 2025
Zero-Shot Video Semantic Segmentation

Zero-Shot Video Semantic Segmentation based on Pre-Trained Diffusion Models

Qian Wang, Abdelrahman Eldesokey, Mohit Mendiratta, Fangneng Zhan, Adam Kortylewski, Christian Theobalt, Peter Wonka

CVPRW 2024
LATENTMAN

LATENTMAN: Generating Consistent Animated Characters using Image Diffusion Models

Abdelrahman Eldesokey, Peter Wonka

CVPR 2021
Uncertainty-Aware CNNs

Uncertainty-Aware CNNs for Depth Completion: Uncertainty from Beginning to End

Abdelrahman Eldesokey, Michael Felsberg, Karl Holmquist, Mikael Persson

Teaching

Current Teaching - KAUST (2023-Present)

Deep Learning

KAUST, MSc level

Generative AI

KAUST, MSc level

Student Supervision

I have supervised more than 20 postgraduate students on various thesis topics in computer vision, deep learning, and generative AI. A selection of supervised theses can be found on my university profile.

Contact

first.last@kaust.edu.sa

KAUST, Saudi Arabia

Abdelrahman Eldesokey

Abdelrahman Eldesokey

Awards and Honors

  • Best Paper Award at VISAPP conference (2021)
  • Affiliation to the Wallenberg AI, Autonomous Systems and Software Program (2017)
  • Graduate Scholarship from Nile University for master's degree (2013)
  • Honor award for outstanding performance in BSc from Mansoura University (2011)
  • One of the Top 10 BSc graduation projects in Egypt by IEEE Gold (2011)