Max Robotics — Global Service Robot Directory + US Certification

About

Alireza Rezazadeh is a researcher at the forefront of robot-assisted rehabilitation and robotic manipulation, whose work bridges the gap between human motor recovery and autonomous machine learning. His early contributions focus on assist-as-needed rehabilitation therapy, where he pioneered a bicycle cranking model that uses learning from demonstration to personalize robotic therapy for elderly and disabled patients—a framework that has garnered 22 citations for its practical approach to restoring motor function. Rezazadeh further advanced co-operative rehabilitation with a Gaussian mixture model for impedance-based tasks, earning 17 citations for enabling robots to adaptively assist patients in activities of daily living. More recently, he has shifted toward unsupervised learning for robotic manipulation, introducing hierarchical graph neural networks for 6D pose estimation of in-hand objects (7 citations) and KINet, an unsupervised forward model for robotic pushing that eliminates the need for ground-truth object labels (5 citations). His latest work, SlotGNN, achieves unsupervised discovery of multi-object representations and visual dynamics, marking a significant step toward robots that learn from raw visual data without human annotation. With a trajectory from clinical rehabilitation to cutting-edge self-supervised learning, Rezazadeh’s research is shaping how robots both assist humans and understand their physical world.