🎓 MSc Robotics & AI — Queensland University of Technology (QUT), Brisbane 🔬 Research tactile-visual imitation learning for robotic manipulation 🤖 Currently: integrating Papillarray tactile sensors into the Universal Manipulation Interface (UMI) 📍 Brisbane, Australia | Open to research & industry roles in robotics / embodied AI
My thesis extends the UMI framework with tactile feedback, enabling the manipulator to learn manipulation policies from human demonstrations that include force and Vision. I achieved <15 ms visual–tactile synchronisation error and validated the pipeline on contact-rich tasks where vision alone is insufficient.
Supervised by Prof. Niko Suenderhauf · QUT Centre for Robotics
Python ROS 2 PyTorch OpenCV C++ Linux Git
🦾 UMI + Tactile Sensing — Visual-tactile sync for imitation learning · <15 ms error 👁️ Stereo Visual Odometry — SuperPoint + LightGlue + PnP RANSAC · 0.75% RTE on KITTI ✋ RGB-D Grasp Detection — Mask R-CNN + GG-CNN two-stage pipeline
💼 LinkedIn: www.linkedin.com/in/omar-mohamed-804b48201
🌐 Portfolio:https://Omar-Mohamed5723.github.io/
📧 Email: omaralbadry7572@gmail.com