Physics-informed Variational Autoencoders for Improved Robustness to Environmental Factors of Variation

TL;DR

This paper presents p$^3$VAE, a physics-informed variational autoencoder that incorporates physical knowledge into the latent space, enhancing robustness, interpretability, and disentanglement in data representations under varying environmental conditions.

Contribution

Introduction of p$^3$VAE, a semi-supervised framework combining neural networks with physics layers to improve data representation robustness and interpretability.

Findings

p$^3$VAE outperforms traditional models in extrapolation tasks.

The model demonstrates effective disentanglement of physical factors.

Experiments show improved robustness to environmental variations.

Abstract

The combination of machine learning models with physical models is a recent research path to learn robust data representations. In this paper, we introduce p$^3$VAE, a variational autoencoder that integrates prior physical knowledge about the latent factors of variation that are related to the data acquisition conditions. p$^3$VAE combines standard neural network layers with non-trainable physics layers in order to partially ground the latent space to physical variables. We introduce a semi-supervised learning algorithm that strikes a balance between the machine learning part and the physics part. Experiments on simulated and real data sets demonstrate the benefits of our framework against competing physics-informed and conventional machine learning models, in terms of extrapolation capabilities and interpretability. In particular, we show that p$^3$VAE naturally has interesting disentanglement capabilities. Our code and data have been made publicly available at https://github.com/Romain3Ch216/p3VAE.