Epsilon Consistent Mixup: An Adaptive Consistency-Interpolation Tradeoff

• URL: http://arxiv.org/abs/2104.09452v1
• Date: Mon, 19 Apr 2021 17:10:31 GMT
• Title: Epsilon Consistent Mixup: An Adaptive Consistency-Interpolation Tradeoff
• Authors: Vincent Pisztora, Yanglan Ou, Xiaolei Huang, Francesca Chiaromonte, Jia Li
• Abstract summary: $epsilon$mu is a data-based structural regularization technique that combines Mixup's linear with consistency regularization in the Mixup direction. It is shown to improve semi-supervised classification accuracy on the SVHN and CIFAR10 benchmark datasets. In particular, $epsilon$mu is found to produce more accurate synthetic labels and more confident predictions than Mixup.
• Score: 19.03167022268852
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: In this paper we propose $\epsilon$-Consistent Mixup ($\epsilon$mu). $\epsilon$mu is a data-based structural regularization technique that combines Mixup's linear interpolation with consistency regularization in the Mixup direction, by compelling a simple adaptive tradeoff between the two. This learnable combination of consistency and interpolation induces a more flexible structure on the evolution of the response across the feature space and is shown to improve semi-supervised classification accuracy on the SVHN and CIFAR10 benchmark datasets, yielding the largest gains in the most challenging low label-availability scenarios. Empirical studies comparing $\epsilon$mu and Mixup are presented and provide insight into the mechanisms behind $\epsilon$mu's effectiveness. In particular, $\epsilon$mu is found to produce more accurate synthetic labels and more confident predictions than Mixup.

Related papers

Err

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.