Saved in:
Bibliographic Details
Main Authors: Shen, Lifeng, Yu, Jincheng, Yang, Hansi, Kwok, James T.
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2406.01417
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917683284410368
author Shen, Lifeng
Yu, Jincheng
Yang, Hansi
Kwok, James T.
author_facet Shen, Lifeng
Yu, Jincheng
Yang, Hansi
Kwok, James T.
contents Mixup and its variants form a popular class of data augmentation techniques.Using a random sample pair, it generates a new sample by linear interpolation of the inputs and labels. However, generating only one single interpolation may limit its augmentation ability. In this paper, we propose a simple yet effective extension called multi-mix, which generates multiple interpolations from a sample pair. With an ordered sequence of generated samples, multi-mix can better guide the training process than standard mixup. Moreover, theoretically, this can also reduce the stochastic gradient variance. Extensive experiments on a number of synthetic and large-scale data sets demonstrate that multi-mix outperforms various mixup variants and non-mixup-based baselines in terms of generalization, robustness, and calibration.
format Preprint
id arxiv_https___arxiv_org_abs_2406_01417
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Mixup Augmentation with Multiple Interpolations
Shen, Lifeng
Yu, Jincheng
Yang, Hansi
Kwok, James T.
Machine Learning
Computer Vision and Pattern Recognition
Mixup and its variants form a popular class of data augmentation techniques.Using a random sample pair, it generates a new sample by linear interpolation of the inputs and labels. However, generating only one single interpolation may limit its augmentation ability. In this paper, we propose a simple yet effective extension called multi-mix, which generates multiple interpolations from a sample pair. With an ordered sequence of generated samples, multi-mix can better guide the training process than standard mixup. Moreover, theoretically, this can also reduce the stochastic gradient variance. Extensive experiments on a number of synthetic and large-scale data sets demonstrate that multi-mix outperforms various mixup variants and non-mixup-based baselines in terms of generalization, robustness, and calibration.
title Mixup Augmentation with Multiple Interpolations
topic Machine Learning
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2406.01417