Generalized Kullback-Leibler Divergence Loss

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Cui, Jiequan, Zhu, Beier, Xu, Qingshan, Tian, Zhuotao, Qi, Xiaojuan, Yu, Bei, Zhang, Hanwang, Hong, Richang
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866929752782143488
author Cui, Jiequan
Zhu, Beier
Xu, Qingshan
Tian, Zhuotao
Qi, Xiaojuan
Yu, Bei
Zhang, Hanwang
Hong, Richang
author_facet Cui, Jiequan
Zhu, Beier
Xu, Qingshan
Tian, Zhuotao
Qi, Xiaojuan
Yu, Bei
Zhang, Hanwang
Hong, Richang
contents In this paper, we delve deeper into the Kullback-Leibler (KL) Divergence loss and mathematically prove that it is equivalent to the Decoupled Kullback-Leibler (DKL) Divergence loss that consists of (1) a weighted Mean Square Error (wMSE) loss and (2) a Cross-Entropy loss incorporating soft labels. Thanks to the decoupled structure of DKL loss, we have identified two areas for improvement. Firstly, we address the limitation of KL loss in scenarios like knowledge distillation by breaking its asymmetric optimization property along with a smoother weight function. This modification effectively alleviates convergence challenges in optimization, particularly for classes with high predicted scores in soft labels. Secondly, we introduce class-wise global information into KL/DKL to reduce bias arising from individual samples. With these two enhancements, we derive the Generalized Kullback-Leibler (GKL) Divergence loss and evaluate its effectiveness by conducting experiments on CIFAR-10/100, ImageNet, and vision-language datasets, focusing on adversarial training, and knowledge distillation tasks. Specifically, we achieve new state-of-the-art adversarial robustness on the public leaderboard -- RobustBench and competitive knowledge distillation performance across CIFAR/ImageNet models and CLIP models, demonstrating the substantial practical merits. Our code is available at https://github.com/jiequancui/DKL.
format Preprint
id arxiv_https___arxiv_org_abs_2503_08038
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Generalized Kullback-Leibler Divergence Loss
Cui, Jiequan
Zhu, Beier
Xu, Qingshan
Tian, Zhuotao
Qi, Xiaojuan
Yu, Bei
Zhang, Hanwang
Hong, Richang
Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
In this paper, we delve deeper into the Kullback-Leibler (KL) Divergence loss and mathematically prove that it is equivalent to the Decoupled Kullback-Leibler (DKL) Divergence loss that consists of (1) a weighted Mean Square Error (wMSE) loss and (2) a Cross-Entropy loss incorporating soft labels. Thanks to the decoupled structure of DKL loss, we have identified two areas for improvement. Firstly, we address the limitation of KL loss in scenarios like knowledge distillation by breaking its asymmetric optimization property along with a smoother weight function. This modification effectively alleviates convergence challenges in optimization, particularly for classes with high predicted scores in soft labels. Secondly, we introduce class-wise global information into KL/DKL to reduce bias arising from individual samples. With these two enhancements, we derive the Generalized Kullback-Leibler (GKL) Divergence loss and evaluate its effectiveness by conducting experiments on CIFAR-10/100, ImageNet, and vision-language datasets, focusing on adversarial training, and knowledge distillation tasks. Specifically, we achieve new state-of-the-art adversarial robustness on the public leaderboard -- RobustBench and competitive knowledge distillation performance across CIFAR/ImageNet models and CLIP models, demonstrating the substantial practical merits. Our code is available at https://github.com/jiequancui/DKL.
title Generalized Kullback-Leibler Divergence Loss
topic Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2503.08038