Saved in:
Bibliographic Details
Main Authors: Shen, Yan, Ji, Zhanghexuan, Ma, Chunwei, Gao, Mingchen
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2402.03588
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909094473891840
author Shen, Yan
Ji, Zhanghexuan
Ma, Chunwei
Gao, Mingchen
author_facet Shen, Yan
Ji, Zhanghexuan
Ma, Chunwei
Gao, Mingchen
contents Domain adversarial adaptation in a continual setting poses a significant challenge due to the limitations on accessing previous source domain data. Despite extensive research in continual learning, the task of adversarial adaptation cannot be effectively accomplished using only a small number of stored source domain data, which is a standard setting in memory replay approaches. This limitation arises from the erroneous empirical estimation of $\gH$-divergence with few source domain samples. To tackle this problem, we propose a double-head discriminator algorithm, by introducing an addition source-only domain discriminator that are trained solely on source learning phase. We prove that with the introduction of a pre-trained source-only domain discriminator, the empirical estimation error of $\gH$-divergence related adversarial loss is reduced from the source domain side. Further experiments on existing domain adaptation benchmark show that our proposed algorithm achieves more than 2$\%$ improvement on all categories of target domain adaptation task while significantly mitigating the forgetting on source domain.
format Preprint
id arxiv_https___arxiv_org_abs_2402_03588
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Continual Domain Adversarial Adaptation via Double-Head Discriminators
Shen, Yan
Ji, Zhanghexuan
Ma, Chunwei
Gao, Mingchen
Machine Learning
Artificial Intelligence
Domain adversarial adaptation in a continual setting poses a significant challenge due to the limitations on accessing previous source domain data. Despite extensive research in continual learning, the task of adversarial adaptation cannot be effectively accomplished using only a small number of stored source domain data, which is a standard setting in memory replay approaches. This limitation arises from the erroneous empirical estimation of $\gH$-divergence with few source domain samples. To tackle this problem, we propose a double-head discriminator algorithm, by introducing an addition source-only domain discriminator that are trained solely on source learning phase. We prove that with the introduction of a pre-trained source-only domain discriminator, the empirical estimation error of $\gH$-divergence related adversarial loss is reduced from the source domain side. Further experiments on existing domain adaptation benchmark show that our proposed algorithm achieves more than 2$\%$ improvement on all categories of target domain adaptation task while significantly mitigating the forgetting on source domain.
title Continual Domain Adversarial Adaptation via Double-Head Discriminators
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2402.03588