Seyed Kamyar Seyed Ghasemipour

About

Seyed Kamyar Seyed Ghasemipour is a leading researcher at the intersection of reinforcement learning and dexterous robotic manipulation, with a particular focus on advancing bi-manual and multi-arm robotic systems. His work addresses the critical limitations of single-arm robots, which are largely confined to simple pick-and-place tasks, by developing scalable sim-to-real transfer techniques for complex, dual-arm coordination. His most impactful contributions include pioneering methods for bi-manual manipulation and attachment, as well as large-scale structured reinforcement learning for multi-part assembly—demonstrated in environments like "Blocks Assemble!" where agents learn to connect magnet blocks. With over 11 citations across his top papers, Ghasemipour’s research has significantly expanded the range of solvable robotic tasks, from object rearrangement to intricate assembly operations. His achievements include successfully transferring learned policies from simulation to real-world hardware, bridging the sim-to-real gap for dual-arm platforms. This work not only advances autonomous robotics but also serves as a diagnostic tool for training embodied intelligent agents, positioning Ghasemipour as a key figure in the push toward more versatile and capable robotic systems.

Research Focus

Key Achievements

Top Papers

1. 1
   Bi-Manual Manipulation and Attachment via Sim-to-Real Reinforcement\n Learning

   5 citations · 2022
2. 2
   Blocks Assemble! Learning to Assemble with Large-Scale Structured Reinforcement Learning

   4 citations · 2022
3. 3
   Bi-Manual Block Assembly via Sim-to-Real Reinforcement Learning

   2 citations · 2023

Key Collaborators

Contact & Links