Contact-invariant optimization for hand manipulation

By Igor Mordatch, Zoran Popovic, E. Todorov. 2012

237 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Solving Rubik's Cube with a Robot Hand (2019, neutral): “[70] and Bai et al.”
Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations (2017, methodology): “a) Model-based trajectory optimization: : Modelbased trajectory optimization methods [31], [24], [38] have demonstrated impressive results in simulated domains, particularly when the dynamics can be adjusted or relaxed to make them more tractable (as, e.”
Dexterous Manipulation with Deep Reinforcement Learning: Efficient, General, and Low-Cost (2018, background): “In simulation, [10], [11], [12] have shown the ability to synthesize complex behaviors when provided with a ground truth model.”
Learning dexterous in-hand manipulation (2018, neutral): “Mordatch et al. [40] and Bai et al. [5] propose methods to generate trajectories for complex and dynamic in-hand manipulation, but results are limited to simulation.”
Deep Dynamics Models for Learning Dexterous Manipulation (2019, background): “In fact, methods that assume perfect knowledge of system dynamics can achieve very impressive manipulation behaviors [15, 16] using generally applicable learning and control techniques.”
Learning dexterous manipulation for a soft robotic hand from human demonstrations (2016, methodology): “Optimization-based techniques [8] have also been used for in-hand manipulation tasks.”
Robust trajectory optimization under frictional contact with iterative learning (2015, methodology): “Some numerical approaches have been presented for dexterous manipulation (Liu 2009; Mordatch et al. 2012) as well as non-prehensile grasp (Lertkultanon and Pham 2014; Lynch and Mason 1996; Pham et al. 2014), but extending them to high dimensional problems has proven computationally challenging.”
Learning Dexterous In-Hand Manipulation. (2018, unknown): “[40] and Bai et al.”
Multi-Agent Manipulation via Locomotion using Hierarchical Sim2Real (2019, background): “Multi-Agent Manipulation via Locomotion The dual relationship between manipulation and locomotion [14, 15] is evident in several previous works using legged robots for object manipulation [16].”
Robot Parkour Learning (2023, background): “This idea has also been explored in robot manipulation [116, 117].”
Learning to Solve a Rubik’s Cube with a Dexterous Hand (2019, background): “Traditional approaches [1], [2] studied simple tasks including grasping and picking up objects in simulation.”
Physically consistent state estimation and system identification for contacts (2015, background): “Also, that idea can help with control for hand manipulation [18].”
Logic-Geometric Programming: An Optimization-Based Approach to Combined Task and Motion Planning (2015, methodology): “[Mordatch et al., 2012] presented an impressive method for optimizing full manipulation trajectories, based on a contactinvariant optimization approach that originated in locomotion research.”
ContactNets: Learning of Discontinuous Contact Dynamics with Smooth, Implicit Representations (2020, background): “While recent accomplishments in planning [1, 2] and control [3], suggest that a model of the robot-environment system’s dynamics is a highly useful formalism for capturing this behavior, producing an accurate model from data is a challenging task for manipulation systems due to the complex behaviors”
A direct method for trajectory optimization of rigid bodies through contact (2014, background): “Broadly speaking, compared to this paper, the work in [14, 25, 24] make varying relaxations of physical or dynamic constraints to pose more tractable (and unconstrained) optimization problems.”
Real-time behaviour synthesis for dynamic hand-manipulation (2014, unknown): “Contact invariant trajectory optimization [11] is an offline method capable of synthesising hand manipulation but its slow and trades physics for visually expressive manipulation behavior.”
Fast, strong and compliant pneumatic actuation for dexterous tendon-driven hands (2013, background): “For a non operational emergency, circuit breakers are provided at all levels from top, which powers down the entire system, to bottom level, which powers down individual subcomponents.”
Grasp'D: Differentiable Contact-rich Grasp Synthesis for Multi-fingered Hands (2022, methodology): “Inspired by Contact-Invariant Optimization [64, 65] we relax the problem making it more forgiving to gradient-based optimization.”