Disentangled State Space Models: Unsupervised Learning of Dynamics Across Heterogeneous Environments
Abstract
Sequential data often originates from diverse environments. Across them exist both shared regularities and environment specifics. To learn robust cross-environment descriptions of sequences we introduce disentangled state space models (DSSM). In the latent space of DSSM environment-invariant state dynamics is explicitly disentangled from environment-specific information governing that dynamics. We empirically show that such separation enables robust prediction, sequence manipulation and environment characterization. We also propose an unsupervised VAE-based training procedure to learn DSSM as Bayesian filters. In our experiments, we demonstrate state-of-the-art performance in controlled generation and prediction of bouncing ball video sequences across varying gravitational influences.
Cite
Text
Miladinović et al. "Disentangled State Space Models: Unsupervised Learning of Dynamics Across Heterogeneous Environments." ICLR 2019 Workshops: DeepGenStruct, 2019.Markdown
[Miladinović et al. "Disentangled State Space Models: Unsupervised Learning of Dynamics Across Heterogeneous Environments." ICLR 2019 Workshops: DeepGenStruct, 2019.](https://mlanthology.org/iclrw/2019/miladinovic2019iclrw-disentangled/)BibTeX
@inproceedings{miladinovic2019iclrw-disentangled,
title = {{Disentangled State Space Models: Unsupervised Learning of Dynamics Across Heterogeneous Environments}},
author = {Miladinović, Ðorđe and Gondal, Waleed and Schölkopf, Bernhard and Buhmann, Joachim M. and Bauer, Stefan},
booktitle = {ICLR 2019 Workshops: DeepGenStruct},
year = {2019},
url = {https://mlanthology.org/iclrw/2019/miladinovic2019iclrw-disentangled/}
}