Towards Combating Frequency Simplicity-biased Learning for Domain Generalization

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: He, Xilin, Hu, Jingyu, Lin, Qinliang, Luo, Cheng, Xie, Weicheng, Song, Siyang, Khan, Muhammad Haris, Shen, Linlin
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912079878815744
author He, Xilin
Hu, Jingyu
Lin, Qinliang
Luo, Cheng
Xie, Weicheng
Song, Siyang
Khan, Muhammad Haris
Shen, Linlin
author_facet He, Xilin
Hu, Jingyu
Lin, Qinliang
Luo, Cheng
Xie, Weicheng
Song, Siyang
Khan, Muhammad Haris
Shen, Linlin
contents Domain generalization methods aim to learn transferable knowledge from source domains that can generalize well to unseen target domains. Recent studies show that neural networks frequently suffer from a simplicity-biased learning behavior which leads to over-reliance on specific frequency sets, namely as frequency shortcuts, instead of semantic information, resulting in poor generalization performance. Despite previous data augmentation techniques successfully enhancing generalization performances, they intend to apply more frequency shortcuts, thereby causing hallucinations of generalization improvement. In this paper, we aim to prevent such learning behavior of applying frequency shortcuts from a data-driven perspective. Given the theoretical justification of models' biased learning behavior on different spatial frequency components, which is based on the dataset frequency properties, we argue that the learning behavior on various frequency components could be manipulated by changing the dataset statistical structure in the Fourier domain. Intuitively, as frequency shortcuts are hidden in the dominant and highly dependent frequencies of dataset structure, dynamically perturbating the over-reliance frequency components could prevent the application of frequency shortcuts. To this end, we propose two effective data augmentation modules designed to collaboratively and adaptively adjust the frequency characteristic of the dataset, aiming to dynamically influence the learning behavior of the model and ultimately serving as a strategy to mitigate shortcut learning. Code is available at AdvFrequency (https://github.com/C0notSilly/AdvFrequency).
format Preprint
id arxiv_https___arxiv_org_abs_2410_16146
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Towards Combating Frequency Simplicity-biased Learning for Domain Generalization
He, Xilin
Hu, Jingyu
Lin, Qinliang
Luo, Cheng
Xie, Weicheng
Song, Siyang
Khan, Muhammad Haris
Shen, Linlin
Machine Learning
Computer Vision and Pattern Recognition
Domain generalization methods aim to learn transferable knowledge from source domains that can generalize well to unseen target domains. Recent studies show that neural networks frequently suffer from a simplicity-biased learning behavior which leads to over-reliance on specific frequency sets, namely as frequency shortcuts, instead of semantic information, resulting in poor generalization performance. Despite previous data augmentation techniques successfully enhancing generalization performances, they intend to apply more frequency shortcuts, thereby causing hallucinations of generalization improvement. In this paper, we aim to prevent such learning behavior of applying frequency shortcuts from a data-driven perspective. Given the theoretical justification of models' biased learning behavior on different spatial frequency components, which is based on the dataset frequency properties, we argue that the learning behavior on various frequency components could be manipulated by changing the dataset statistical structure in the Fourier domain. Intuitively, as frequency shortcuts are hidden in the dominant and highly dependent frequencies of dataset structure, dynamically perturbating the over-reliance frequency components could prevent the application of frequency shortcuts. To this end, we propose two effective data augmentation modules designed to collaboratively and adaptively adjust the frequency characteristic of the dataset, aiming to dynamically influence the learning behavior of the model and ultimately serving as a strategy to mitigate shortcut learning. Code is available at AdvFrequency (https://github.com/C0notSilly/AdvFrequency).
title Towards Combating Frequency Simplicity-biased Learning for Domain Generalization
topic Machine Learning
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2410.16146