Ivan Laptev
Centre National de la Recherche Scientifique, Département d'Informatique, Université Paris Sciences et Lettres, Group Image (Poland), Centre de Recherche en Informatique, Institut national de recherche en sciences et technologies du numérique, Mohamed bin Zayed University of Artificial Intelligence, Technical University of Munich, École Normale Supérieure
Papers
27
Total Citations
456
H-Index
12
About
Ivan Laptev is a prominent computer vision and robotics researcher whose work sits at the dynamic intersection of visual perception, robot learning, and manipulation planning. His research addresses some of the most challenging problems in modern robotics: enabling machines to understand and interact with the physical world through visual inputs alone. Laptev's most influential contributions span visually guided rearrangement planning, where his Monte-Carlo Tree Search approach (73 citations) demonstrated how robots can efficiently plan object manipulation from raw RGB inputs. His work on sim-to-real transfer (45 citations) tackles the fundamental challenge of bridging the gap between simulated training environments and real-world deployment—a critical bottleneck in practical robotics. He has also made significant strides in differentiable physics simulation (44 citations) and contact modeling (36 citations), advancing the mathematical foundations that underpin modern robot control pipelines. Particularly noteworthy is his research into learning from human demonstrations, including 3D hand-object reconstruction from monocular video (57 citations) and video-conditioned policy learning, which moves robotics toward systems that can learn new tasks simply by watching people. Collectively, Laptev's body of work reflects a coherent vision: building robots that perceive, reason, and act with human-like flexibility in unstructured environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2Towards Unconstrained Joint Hand-Object Reconstruction From RGB Videos57 citations · 2021
- 3Learning to Augment Synthetic Images for Sim2Real Policy Transfer45 citations · 2019
- 4Differentiable Simulation for Physical System Identification44 citations · 2021
- 5
- 6Contact Models in Robotics: A Comparative Analysis36 citations · 2024
- 7Goal-Conditioned Reinforcement Learning with Imagined Subgoals26 citations · 2021
- 8Learning Video-Conditioned Policies for Unseen Manipulation Tasks14 citations · 2023
- 9Editorial- Deep Learning for Computer Vision13 citations · 2017
- 10