Non-linear State-space Model Identification from Video Data using Deep Encoders

  title={Non-linear State-space Model Identification from Video Data using Deep Encoders},
  author={Gerben Izaak Beintema and Roland T{\'o}th and Maarten Schoukens},

Figures and Tables from this paper

On compression rate of quantum autoencoders: Control design, numerical and experimental realization
The upper bound of the compression rate is theoretically proven using eigen-decomposition and matrix differentiation, which is determined by the eigenvalues of the density matrix representation of the input states.
Deep Identification of Nonlinear Systems in Koopman Form
The present paper treats the identification of nonlinear dynamical systems using Koopman-based deep state-space encoders using the the deepSI toolbox in Python to lower the computational need of the simulation error-based training.


Universal approximation bounds for superpositions of a sigmoidal function
  • A. Barron
  • Computer Science
    IEEE Trans. Inf. Theory
  • 1993
The approximation rate and the parsimony of the parameterization of the networks are shown to be advantageous in high-dimensional settings and the integrated squared approximation error cannot be made smaller than order 1/n/sup 2/d/ uniformly for functions satisfying the same smoothness assumption.
Variational Nonlinear State Estimation.
A solution based on the variational inference principle is developed, which offers the key advantage of a flexible, but principled, mechanism for approximating the required distributions and outperforms alternative assumed density approaches to state estimation.
Model-based Reinforcement Learning: A Survey
A survey of the integration of model-based reinforcement learning and planning, better known as model- based reinforcement learning, and a broad conceptual overview of planning-learning combinations for MDP optimization are presented.
Integrated Neural Networks for Nonlinear Continuous-Time System Identification
This letter introduces a novel neural network architecture, called Integrated Neural Network (INN), for direct identification of nonlinear continuous-time dynamical models in state-space representation and its effectiveness is assessed against the Cascaded Tanks System benchmark.
The trade-off between long-term memory and smoothness for recurrent networks
This work provides insight into the trade-off between the smoothness of the cost function and the memory retention capabilities of the network in terms of the Lipschitz constant of the dynamics modeled by the network.
Learning Nonlinear State-Space Models Using Deep Autoencoders
This work introduces a new methodology for the identification of nonlinear state-space models using machine-learning techniques based on deep autoencoders for dimensionality reduction and neural networks and shows its capability of fitting a nonlinear model from an input/output dataset generated by a benchmark nonlinear system.
Iterative Excitation Signal Design for Nonlinear Dynamic Black-Box Models
Subspace Identification for Linear Systems: Theory ― Implementation ― Applications
This book focuses on the theory, implementation and applications of subspace identification algorithms for linear time-invariant finitedimensional dynamical systems, which allow for a fast, straightforward and accurate determination of linear multivariable models from measured inputoutput data.