Rethinking Layer-wise Feature Amounts in Convolutional Neural Network Architectures
Machine Learning
2018-12-17 v1 Computer Vision and Pattern Recognition
Machine Learning
Abstract
We characterize convolutional neural networks with respect to the relative amount of features per layer. Using a skew normal distribution as a parametrized framework, we investigate the common assumption of monotonously increasing feature-counts with higher layers of architecture designs. Our evaluation on models with VGG-type layers on the MNIST, Fashion-MNIST and CIFAR-10 image classification benchmarks provides evidence that motivates rethinking of our common assumption: architectures that favor larger early layers seem to yield better accuracy.
Cite
@article{arxiv.1812.05836,
title = {Rethinking Layer-wise Feature Amounts in Convolutional Neural Network Architectures},
author = {Martin Mundt and Sagnik Majumder and Tobias Weis and Visvanathan Ramesh},
journal= {arXiv preprint arXiv:1812.05836},
year = {2018}
}
Comments
Accepted at the Critiquing and Correcting Trends in Machine Learning (CRACT) Workshop at the 32nd Conference on Neural Information Processing Systems (NeurIPS 2018)