Behavior-1k: A benchmark for embodied ai with 1,000 everyday activities and realistic simulation

C Li, R Zhang, J Wong, C Gokmen… - … on Robot Learning, 2023 - proceedings.mlr.press
We present BEHAVIOR-1K, a comprehensive simulation benchmark for human-centered
robotics. BEHAVIOR-1K includes two components, guided and motivated by the results of an …

Grasp'd: Differentiable contact-rich grasp synthesis for multi-fingered hands

D Turpin, L Wang, E Heiden, YC Chen… - … on Computer Vision, 2022 - Springer
The study of hand-object interaction requires generating viable grasp poses for high-
dimensional multi-finger models, often relying on analytic grasp synthesis which tends to …

A review of differentiable simulators

R Newbury, J Collins, K He, J Pan, I Posner… - IEEE …, 2024 - ieeexplore.ieee.org
Differentiable simulators continue to push the state of the art across a range of domains
including computational physics, robotics, and machine learning. Their main value is the …

Diffskill: Skill abstraction from differentiable physics for deformable object manipulations with tools

X Lin, Z Huang, Y Li, JB Tenenbaum, D Held… - arxiv preprint arxiv …, 2022 - arxiv.org
We consider the problem of sequential robotic manipulation of deformable objects using
tools. Previous works have shown that differentiable physics simulators provide gradients to …

Rethinking optimization with differentiable simulation from a global perspective

R Antonova, J Yang… - Conference on Robot …, 2023 - proceedings.mlr.press
Differentiable simulation is a promising toolkit for fast gradient-based policy optimization and
system identification. However, existing approaches to differentiable simulation have largely …

Planning with spatial-temporal abstraction from point clouds for deformable object manipulation

X Lin, C Qi, Y Zhang, Z Huang, K Fragkiadaki… - arxiv preprint arxiv …, 2022 - arxiv.org
Effective planning of long-horizon deformable object manipulation requires suitable
abstractions at both the spatial and temporal levels. Previous methods typically either focus …