Max Robotics — Global Service Robot Directory + US Certification

About

Reza Mahjourian is a robotics and computer vision researcher whose work spans unsupervised depth estimation, robot learning, and real-world autonomous systems. He is perhaps best known for his 2019 paper "Depth Prediction without the Sensors," which garnered over 490 citations and introduced an influential framework for learning scene depth and ego-motion from monocular video without explicit sensor supervision — a breakthrough that significantly advanced affordable robot perception and autonomous driving research. Beyond computer vision, Mahjourian has made substantial contributions to robotic motor skills, particularly in the domain of high-speed robot table tennis. His hierarchical policy design work demonstrated sample-efficient learning through self-play, laying groundwork for increasingly capable physical agents. This trajectory culminated in his landmark achievement of the first learned robotic agent to reach amateur human-level performance in competitive table tennis — a milestone celebrated across the robotics community. His 2023 deep-dive into the full robotic table tennis stack further showcased his systems-level thinking, integrating perception, planning, and control into a cohesive, deployable platform. Mahjourian's research consistently bridges foundational machine learning with real-world robotic deployment, making him a notable figure for students interested in embodied AI, autonomous perception, and the frontier of human-competitive robot performance.