Baruch Tabanpour
Baruch Tabanpour is a co-author of MuJoCo Playground and its 2025 RSS demonstration paper. At the time of publication, Tabanpour was affiliated with Google DeepMind. Playground is an open-source robot learning framework built with MuJoCo MJX. It combines physics, batch rendering, and task environments for locomotion and manipulation. The paper reports robot deployments using policies trained from state and image inputs.
- Works
- 8
- Citations
- 56
- h-index
- 4
Publication numbers from OpenAlex, as of 14 October 2026.
Timeline
-
2025
An open robot learning framework
Tabanpour coauthored MuJoCo Playground, published as a manuscript in February 2025. The paper describes an open-source framework built with MuJoCo MJX. Its aim is to connect simulation, policy training, and transfer to physical robots in a common software stack.
-
2025
Locomotion and manipulation environments
The framework includes tasks for quadrupeds, humanoids, dexterous hands, and robot arms. The paper describes environments that use physics simulation to train action policies. These are capabilities of the shared project.
-
2025
Training with state and images
The paper describes a stack with a physics engine, batch rendering, and training environments. It reports transfer of policies trained from state inputs and pixel inputs. Those demonstrations apply to the robots and conditions in the study; they do not guarantee transfer for every platform.
-
2025
A conference demonstration
Tabanpour also coauthored Demonstrating MuJoCo Playground at RSS 2025. Tabanpour was affiliated with Google DeepMind at the time of publication.
Sources and credits
Photo: We have no published photo of Baruch Tabanpour yet. Send us a free-licensed photo.