Adham Elarabawy

I work on vision research at Google DeepMind, with a focus on video understanding, representations, and long-context capabilities. I also enjoy tennis and cooking, sometimes even successfully.

GitHub LinkedIn X YouTube

Selected Projects

• 

  July 2026

  Gen4U: Unifying Video Generation and Understanding via Diffusion

  A framework that repurposes frozen video diffusion models as general-purpose video encoders with a single forward pass.

  Michael King^*, Aravindh Mahendran^*, Matthew Koichi Grimes, Fedor Kitashov, Adham Elarabawy, Pedro Velez, Maks Ovsjanikov, Viorica Pătrăucean·Google DeepMind

• 

  September 2025

  EmbeddingGemma: Powerful and Lightweight Text Representations

  A lightweight, open text embedding model designed for strong multilingual retrieval and on-device applications.

  Large Google DeepMind collaboration·Adham Elarabawy, contributor

• 

  December 2022

  Direct Inversion

  Optimization-Free Text-Driven Real Image Editing with Diffusion Models.

  Adham Elarabawy, Harish Kamath, Samuel Denton

• 

  December 2020

  OpenQuadruped

  An open-source 3D-printed quadrupedal robot dog with custom IK, gait planning, and ROS integration.

  Adham Elarabawy, Maurice Rahme