Publication Type

Conference Proceeding Article

Version

publishedVersion

Publication Date

6-2022

Abstract

Data mixing (e.g., Mixup, Cutmix, ResizeMix) is an essential component for advancing recognition models. In this paper, we focus on studying its effectiveness in the self-supervised setting. By noticing the mixed images that share the same source images are intrinsically related to each other, we hereby propose SDMP, short for Simple Data Mixing Prior, to capture this straightforward yet essential prior, and position such mixed images as additional positive pairs to facilitate self-supervised representation learning. Our experiments verify that the proposed SDMP enables data mixing to help a set of self-supervised learning frameworks (e.g., MoCo) achieve better accuracy and out-of-distribution robustness. More notably, our SDMP is the first method that successfully leverages data mixing to improve (rather than hurt) the performance of Vision Transformers in the self-supervised setting. Code is publicly available at https://github.com/OliverRensu/SDMP.

Keywords

Categorization, Representation learning, Retrieval, Self- & semi- & meta- recognition, Detection

Discipline

Databases and Information Systems | Graphics and Human Computer Interfaces

Research Areas

Information Systems and Management; Intelligent Systems and Optimization

Publication

Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (CVPR)

First Page

14575

Last Page

14584

ISBN

9781665469463

Identifier

10.1109/CVPR52688.2022.01419

Publisher

IEEE

City or Country

Canada

Copyright Owner and License

Authors

Share

COinS