Max Robotics — Global Service Robot Directory + US Certification

About

Bernard Ghanem is a prominent researcher whose work spans autonomous driving, 3D computer vision, neural architecture search, and embodied artificial intelligence. His research addresses some of the most challenging problems at the intersection of deep learning and real-world perception systems. One of his most influential contributions, "Driving Policy Transfer via Modularity and Abstraction" (2018, 122 citations), tackled the high sample complexity of end-to-end autonomous driving by leveraging simulation environments to bridge the gap between synthetic training and real-world deployment — a landmark step in scalable urban driving research. His work on 3D point clouds, including 3DeformRS, advances the certified robustness of models used in safety-critical applications like surgical robotics and self-driving vehicles, addressing growing concerns over adversarial and real-world deformations. His investigations into latency-constrained neural architecture search for point cloud networks further demonstrate his commitment to practical, deployable AI systems. More recently, Ghanem has explored embodied AI through LLM-driven robotic control frameworks and reinforcement learning from video observations. His earlier foundational work on dynamic textures and swarm modeling reflects a career built on rigorous visual modeling. Across his portfolio, Ghanem consistently bridges theoretical innovation with high-stakes applications, making him a significant voice in modern computer vision and robotics research.