AutoMix: Mixup Networks for Sample Interpolation via Cooperative Barycenter Learning

Abstract

This paper proposes new ways of sample mixing by thinking of the process as generation of barycenter in a metric space for data augmentation. First, we present an optimal-transport-based mixup technique to generate Wasserstein barycenter which works well on images with clean background and is empirically shown complementary to existing mixup methods. Then we generalize mixup to an AutoMix technique by using a learnable network to fit barycenter in a cooperative way between the classifier (a.k.a. discriminator) and generator networks. Experimental results on both multi-class and multi-label prediction tasks show the efficacy of our approach, which is also verified in the presence of unseen categories (open set) and noise.

Cite

Text

Zhu et al. "AutoMix: Mixup Networks for Sample Interpolation via Cooperative Barycenter Learning." Proceedings of the European Conference on Computer Vision (ECCV), 2020. doi:10.1007/978-3-030-58607-2_37

Markdown

[Zhu et al. "AutoMix: Mixup Networks for Sample Interpolation via Cooperative Barycenter Learning." Proceedings of the European Conference on Computer Vision (ECCV), 2020.](https://mlanthology.org/eccv/2020/zhu2020eccv-automix/) doi:10.1007/978-3-030-58607-2_37

BibTeX

@inproceedings{zhu2020eccv-automix,
  title     = {{AutoMix: Mixup Networks for Sample Interpolation via Cooperative Barycenter Learning}},
  author    = {Zhu, Jianchao and Shi, Liangliang and Yan, Junchi and Zha, Hongyuan},
  booktitle = {Proceedings of the European Conference on Computer Vision (ECCV)},
  year      = {2020},
  doi       = {10.1007/978-3-030-58607-2_37},
  url       = {https://mlanthology.org/eccv/2020/zhu2020eccv-automix/}
}