Multi-Prompt Alignment for Multi-Source Unsupervised Domain Adaptation

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Chen, Haoran, Han, Xintong, Wu, Zuxuan, Jiang, Yu-Gang
Format: Preprint
Veröffentlicht: 2022
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866916265360097280
author Chen, Haoran
Han, Xintong
Wu, Zuxuan
Jiang, Yu-Gang
author_facet Chen, Haoran
Han, Xintong
Wu, Zuxuan
Jiang, Yu-Gang
contents Most existing methods for unsupervised domain adaptation (UDA) rely on a shared network to extract domain-invariant features. However, when facing multiple source domains, optimizing such a network involves updating the parameters of the entire network, making it both computationally expensive and challenging, particularly when coupled with min-max objectives. Inspired by recent advances in prompt learning that adapts high-capacity models for downstream tasks in a computationally economic way, we introduce Multi-Prompt Alignment (MPA), a simple yet efficient framework for multi-source UDA. Given a source and target domain pair, MPA first trains an individual prompt to minimize the domain gap through a contrastive loss. Then, MPA denoises the learned prompts through an auto-encoding process and aligns them by maximizing the agreement of all the reconstructed prompts. Moreover, we show that the resulting subspace acquired from the auto-encoding process can easily generalize to a streamlined set of target domains, making our method more efficient for practical usage. Extensive experiments show that MPA achieves state-of-the-art results on three popular datasets with an impressive average accuracy of 54.1% on DomainNet.
format Preprint
id arxiv_https___arxiv_org_abs_2209_15210
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle Multi-Prompt Alignment for Multi-Source Unsupervised Domain Adaptation
Chen, Haoran
Han, Xintong
Wu, Zuxuan
Jiang, Yu-Gang
Computer Vision and Pattern Recognition
Most existing methods for unsupervised domain adaptation (UDA) rely on a shared network to extract domain-invariant features. However, when facing multiple source domains, optimizing such a network involves updating the parameters of the entire network, making it both computationally expensive and challenging, particularly when coupled with min-max objectives. Inspired by recent advances in prompt learning that adapts high-capacity models for downstream tasks in a computationally economic way, we introduce Multi-Prompt Alignment (MPA), a simple yet efficient framework for multi-source UDA. Given a source and target domain pair, MPA first trains an individual prompt to minimize the domain gap through a contrastive loss. Then, MPA denoises the learned prompts through an auto-encoding process and aligns them by maximizing the agreement of all the reconstructed prompts. Moreover, we show that the resulting subspace acquired from the auto-encoding process can easily generalize to a streamlined set of target domains, making our method more efficient for practical usage. Extensive experiments show that MPA achieves state-of-the-art results on three popular datasets with an impressive average accuracy of 54.1% on DomainNet.
title Multi-Prompt Alignment for Multi-Source Unsupervised Domain Adaptation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2209.15210