Exchangeable Generative Models with Flow Scans
Abstract
In this work, we develop a new approach to generative density estimation for exchangeable, non-i.i.d. data. The proposed framework, FlowScan, combines invertible flow transformations with a sorted scan to flexibly model the data while preserving exchangeability. Unlike most existing methods, FlowScan exploits the intradependencies within sets to learn both global and local structure. FlowScan represents the first approach that is able to apply sequential methods to exchangeable density estimation without resorting to averaging over all possible permutations. We achieve new state-of-the-art performance on point cloud and image set modeling.
Cite
Text
Bender et al. "Exchangeable Generative Models with Flow Scans." AAAI Conference on Artificial Intelligence, 2020. doi:10.1609/AAAI.V34I06.6562Markdown
[Bender et al. "Exchangeable Generative Models with Flow Scans." AAAI Conference on Artificial Intelligence, 2020.](https://mlanthology.org/aaai/2020/bender2020aaai-exchangeable/) doi:10.1609/AAAI.V34I06.6562BibTeX
@inproceedings{bender2020aaai-exchangeable,
title = {{Exchangeable Generative Models with Flow Scans}},
author = {Bender, Christopher M. and O'Connor, Kevin and Li, Yang and Garcia, Juan Jose and Oliva, Junier and Zaheer, Manzil},
booktitle = {AAAI Conference on Artificial Intelligence},
year = {2020},
pages = {10053-10060},
doi = {10.1609/AAAI.V34I06.6562},
url = {https://mlanthology.org/aaai/2020/bender2020aaai-exchangeable/}
}