• Corpus ID: 232068635

Physics-Integrated Variational Autoencoders for Robust and Interpretable Generative Modeling

  title={Physics-Integrated Variational Autoencoders for Robust and Interpretable Generative Modeling},
  author={Naoya Takeishi and Alexandros Kalousis},
Integrating physics models within machine learning holds considerable promise toward learning robust models with improved interpretability and abilities to extrapolate. In this work, we focus on the integration of incomplete physics models into deep generative models, variational autoencoders (VAEs) in particular. A key technical challenge is to strike a balance between the incomplete physics model and the learned components (i.e., neural nets) of the complete model, in order to ensure that the… 
Robust Hybrid Learning With Expert Augmentation
It is shown why expert augmentation improves generalization and the practical benefits of augmented hybrid models on a set of controlled experiments, modelling dynamical systems described by ordinary and partial differential equations.
Meta-Parameter Selection for Embedding Generation of Latency Spaces in Auto Encoder Analytics
This ansatz allows us to deduce meaningful meta-parameter limits using OPTICS, DBSCAN, UMAP, t-SNE, and k-MEANS to perform first steps of a meaningful visual analysis in the unsupervised case using a vanilla autoencoder on the MNIST and DeepVALVE data sets.


  • P. Alam
  • Composites Engineering: An A–Z Guide
  • 2021
Hamiltonian Generative Networks
This work introduces the Hamiltonian Generative Network (HGN), the first approach capable of consistently learning Hamiltonian dynamics from high-dimensional observations (such as images) without restrictive domain assumptions, and demonstrates how a simple modification of the network architecture turns HGN into a powerful normalising flow model, called Neural Hamiltonian Flow (NHF), that usesHamiltonian dynamics to model expressive densities.
Self-supervised learning with physics-aware neural networks – I
  • Galaxy model fitting. Monthly Notices of the Royal Astronomical Society, 498(3): 3713–3719
  • 2020
Augmenting physical models with deep networks for complex dynamics forecasting
The APHYNITY framework is introduced, a principled approach for augmenting incomplete physical dynamics described by differential equations with deep data-driven models and can efficiently leverage approximate physical models to accurately forecast the evolution of the system and correctly identify relevant physical parameters.
Disentangling Physical Dynamics From Unknown Factors for Unsupervised Video Prediction
  • V. Guen, Nicolas Thome
  • Computer Science
    2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2020
PhyDNet is introduced, a two-branch deep architecture, which explicitly disentangles PDE dynamics from unknown complementary information, and a new recurrent physical cell (PhyCell) is proposed, inspired from data assimilation techniques, for performing PDE-constrained prediction in latent space.
Self-supervised learning with physics-aware neural networks – I. Galaxy model fitting
The Semantic Autoencoder architecture described here shares the robustness of neural networks while including an explicit model of the data, learns in an unsupervised way and estimates, by construction, parameters with direct physical interpretation.
Hamiltonian Neural Networks
Inspiration from Hamiltonian mechanics is drawn to train models that learn and respect exact conservation laws in an unsupervised manner, and this model trains faster and generalizes better than a regular neural network.
Integrating Model-Driven and Data-Driven Methods for Power System Frequency Stability Assessment and Control
This paper discusses feasible integration approaches for the model-driven and data-driven methods based on the existing achievements, and proposes to integrate both methods for the power system online frequency stability assessment and control.
Modeling chemical processes using prior knowledge and neural networks
The approach is shown to require less data for parameter estimation, produce more accurate and consistent predictions, and provide more reliable extrapolation and the results show that prior knowledge enhances the generalization capabilities of a pure neural network model.
Data-Augmented Contact Model for Rigid Body Simulation
This paper introduces a data-augmented contact model that incorporates analytical solutions with observed data to predict the 3D contact impulse which could result in rigid bodies bouncing, sliding or spinning in all directions.