Composing Graphical Models with Neural Networks for Structured Representations and Fast Inference
Abstract
We propose a general modeling and inference framework that combines the complementary strengths of probabilistic graphical models and deep learning methods. Our model family composes latent graphical models with neural network observation likelihoods. For inference, we use recognition networks to produce local evidence potentials, then combine them with the model distribution using efficient message-passing algorithms. All components are trained simultaneously with a single stochastic variational inference objective. We illustrate this framework by automatically segmenting and categorizing mouse behavior from raw depth video, and demonstrate several other example models.
Cite
Text
Johnson et al. "Composing Graphical Models with Neural Networks for Structured Representations and Fast Inference." Neural Information Processing Systems, 2016.Markdown
[Johnson et al. "Composing Graphical Models with Neural Networks for Structured Representations and Fast Inference." Neural Information Processing Systems, 2016.](https://mlanthology.org/neurips/2016/johnson2016neurips-composing/)BibTeX
@inproceedings{johnson2016neurips-composing,
title = {{Composing Graphical Models with Neural Networks for Structured Representations and Fast Inference}},
author = {Johnson, Matthew J and Duvenaud, David K. and Wiltschko, Alex and Adams, Ryan P. and Datta, Sandeep R},
booktitle = {Neural Information Processing Systems},
year = {2016},
pages = {2946-2954},
url = {https://mlanthology.org/neurips/2016/johnson2016neurips-composing/}
}