• Corpus ID: 218900998

Euclideanizing Flows: Diffeomorphic Reduction for Learning Stable Dynamical Systems

@article{Rana2020EuclideanizingFD,
  title={Euclideanizing Flows: Diffeomorphic Reduction for Learning Stable Dynamical Systems},
  author={Muhammad Asif Rana and Anqi Li and Dieter Fox and Byron Boots and Fabio Tozeto Ramos and Nathan D. Ratliff},
  journal={ArXiv},
  year={2020},
  volume={abs/2005.13143}
}
Robotic tasks often require motions with complex geometric structures. We present an approach to learn such motions from a limited number of human demonstrations by exploiting the regularity properties of human motions e.g. stability, smoothness, and boundedness. The complex motions are encoded as rollouts of a stable dynamical system, which, under a change of coordinates defined by a diffeomorphism, is equivalent to a simple, hand-specified dynamical system. As an immediate result of using… 

Figures from this paper

Linearization and Identification of Multiple-Attractors Dynamical System through Laplacian Eigenmaps

A Graph-based spectral clustering method that takes advantage of a velocity-augmented kernel to connect data-points belonging to the same dynamics, while preserving the natural temporal evolution is proposed.

Learning Stable Vector Fields on Lie Groups

This paper presents a novel vector field model that can guarantee most of the properties of previous approaches i.e., stability, smoothness, and reactivity beyond the Euclidean space.

Learning Stable Vector Fields on Smooth Manifolds

  • Computer Science
This paper presents a novel vector field model that can guarantee most of the properties of previous approaches i.e., stability, smoothness, and reactivity beyond the euclidean space.

Towards Coordinated Robot Motions: End-to-End Learning of Motion Policies on Transform Trees

This paper addresses the challenge of learning motion policies to generate motions for execution of such tasks by decomposing a motion policy into multiple subtask policies, whereby each subtask policy dictates a particular subtask behavior.

Learning Stable Robotic Skills on Riemannian Manifolds

The approach leverages a data-e ffi cient procedure to learn a di ff eomorphic transformation that maps simple stable dynamical systems onto complex robotic skills and shows promising results in terms of learning accuracy and task adaptation capabilities.

Consensus-based Normalizing-Flow Control: A Case Study in Learning Dual-Arm Coordination

—We develop two consensus-based learning algo- rithms for multi-robot systems applied on complex tasks in-volving collision constraints and force interactions, such as the cooperative peg-in-hole

Learning High Dimensional Demonstrations Using Laplacian Eigenmaps

A novel methodology to learn a stable robot control law driven by dynamical systems that enables to model highly nonlinear dynamics in high dimension and overcomes alternative techniques in both precision of reconstruction and number of parameters required for the embedding.

Reactive Motion Generation on Learned Riemannian Manifolds

It is argued that Riemannian manifolds may be learned via human demonstrations in which geodesics are natural motion skills, and proposed technique for facilitating on-the-fly end-effector/multiple-limb obstacle avoidance by reshaping the learned manifold using an obstacleaware ambient metric.

Robot Skill Adaptation via Soft Actor-Critic Gaussian Mixture Models

The Soft Actor- Critic Gaussian Mixture Model (SAC-GMM), a novel hybrid approach that learns robot skills through a dynamical system and adapts the learned skills in their own trajectory distribution space through interactions with the environment, is proposed.

KAM Theory Meets Statistical Learning Theory: Hamiltonian Neural Networks with Non-zero Training Loss

This study provides a generalization error bound for Hamiltonian neural networks by deriving an estimate of the covering number of the gradient of the multi-layer perceptron, which is the key ingredient of the model.

References

SHOWING 1-10 OF 28 REFERENCES

Signaling Properties of Hyaluronan Receptors*

The study of murine cardiac cells derived from hyaluronan synthase 2 (HAS2) knockout mice has provided the most convincing evidence for a signaling capability of hyAluronan, and results suggest that hy aluronan signals through Ras to regulate motility are consistent with previous studies showing that exogenous hyaltonanreceptor interactions regulate Ras signaling.

A guide to medical entomology

  • p. Copyright © 2014 SciRes. OPEN ACCESS
  • 1999

Beyond tumorigenesis: cancer stem cells in metastasis

A unified model of cancer progression is proposed in which CSCs assume a central role in both tumorigenesis and metastasis, and this model will lead to novel therapeutic strategies against metastatic cancer.

Learning position and orientation dynamics from demonstrations via contraction analysis

The CDSP algorithm is evaluated on a publicly available dataset and a synthetic dataset, and is shown to have the lowest and comparable average reproduction errors when compared to state-of-the-art imitation learning algorithms.

Learning Reactive Motion Policies in Multiple Task Spaces from Human Demonstrations

This work decomposes a task into multiple subtasks and learns to reproduce the subtasks by learning stable policies by leveraging the RMPflow framework for motion generation, and finds a stable global policy in the configuration space that enables simultaneous execution of various learned subtasks.

Rendiconti del circolo matematico di Palermo

1. After du Bois-Reymond [7] had constructed the example of a continuous function whose Fourier series diverges at apoint Jordan [12] gave a new criterion for convergence of Fourier series which

[Pneumocystis carinii pneumonia in patients with AIDS].