We introduce a new method for speeding up the inference of deep neural networks. It is somewhat inspired by the reduced-order modeling techniques for dynamical systems.The cornerstone of the proposed method is the maximum volume algorithm. We demonstrate efficiency on neural networks pre-trained on different datasets. We show that in many practical cases it is possible to replace convolutional layers with much smaller fully-connected layers with a relatively small drop in accuracy.
@article{arxiv.1910.06995,
title = {Reduced-Order Modeling of Deep Neural Networks},
author = {Julia Gusak and Talgat Daulbaev and Evgeny Ponomarev and Andrzej Cichocki and Ivan Oseledets},
journal= {arXiv preprint arXiv:1910.06995},
year = {2020}
}