Skip to Main content Skip to Navigation
Conference papers

PDE-Driven Spatiotemporal Disentanglement

Abstract : A recent line of work in the machine learning community addresses the problem of predicting high-dimensional spatiotemporal phenomena by leveraging specific tools from the differential equations theory. Following this direction, we propose in this article a novel and general paradigm for this task based on a resolution method for partial differential equations: the separation of variables. This inspiration allows us to introduce a dynamical interpretation of spatiotemporal disentanglement. It induces a principled model based on learning disentangled spatial and temporal representations of a phenomenon to accurately predict future observations. We experimentally demonstrate the performance and broad applicability of our method against prior state-of-the-art models on physical and synthetic video datasets.
Complete list of metadata
Contributor : Jean-Yves Franceschi Connect in order to contact the contributor
Submitted on : Wednesday, March 17, 2021 - 11:23:21 PM
Last modification on : Wednesday, March 16, 2022 - 3:50:23 AM


Files produced by the author(s)


Distributed under a Creative Commons Attribution 4.0 International License


  • HAL Id : hal-02911067, version 3
  • ARXIV : 2008.01352



Jérémie Donà, Jean-Yves Franceschi, Sylvain Lamprier, Patrick Gallinari. PDE-Driven Spatiotemporal Disentanglement. The Ninth International Conference on Learning Representations, International Conference on Representation Learning, May 2021, Vienne, Austria. ⟨hal-02911067v3⟩



Record views


Files downloads