S. M. Ali Eslami

Papers

Total Citations

H-Index

About

S. M. Ali Eslami is a leading researcher at the intersection of computer vision, generative modeling, and robotics, with a focus on enabling machines to understand and interact with the physical world. His major contributions span two transformative areas: learning-based 3D content generation and real-world robot learning. Eslami is best known for pioneering PolyGen (2020, 65 citations), an autoregressive generative model that directly outputs 3D polygon meshes—a breakthrough that sidesteps the limitations of voxel or point cloud representations, offering a more efficient and geometrically faithful approach to 3D shape synthesis. This work has become foundational for applications in computer graphics, robotics, and game development. More recently, his research on "Towards Real Robot Learning in the Wild" (2021, 5 citations) tackles the challenge of deploying bipedal locomotion algorithms outside controlled lab settings, addressing critical issues of robustness and adaptability. Eslami's work is characterized by its ambition to bridge the gap between theoretical generative models and practical, embodied AI systems, making him a key figure in the push toward machines that can both imagine and act in complex 3D environments.