Physics perception in sloshing scenes with guaranteed thermodynamic consistency
This work addresses physics perception challenges in fluid dynamics for applications like augmented reality, though it is incremental as it builds on existing reduced-order modeling and neural network techniques.
The authors tackled the problem of reconstructing the full state of sloshing liquids from limited free-surface measurements by using recurrent neural networks with thermodynamic inductive biases, resulting in a system that predicts future fluid states in real-time with real-world computer vision integration.
Physics perception very often faces the problem that only limited data or partial measurements on the scene are available. In this work, we propose a strategy to learn the full state of sloshing liquids from measurements of the free surface. Our approach is based on recurrent neural networks (RNN) that project the limited information available to a reduced-order manifold so as to not only reconstruct the unknown information, but also to be capable of performing fluid reasoning about future scenarios in real time. To obtain physically consistent predictions, we train deep neural networks on the reduced-order manifold that, through the employ of inductive biases, ensure the fulfillment of the principles of thermodynamics. RNNs learn from history the required hidden information to correlate the limited information with the latent space where the simulation occurs. Finally, a decoder returns data back to the high-dimensional manifold, so as to provide the user with insightful information in the form of augmented reality. This algorithm is connected to a computer vision system to test the performance of the proposed methodology with real information, resulting in a system capable of understanding and predicting future states of the observed fluid in real-time.