Abstract:Domain adaptation enhances generalizability of a model across domains with domain shifts. Most research effort has been spent on Unsupervised Domain Adaption (UDA) which trains a model jointly with labeled source data and unlabeled target data. This paper studies how much it can help address domain shifts if we further have a few target samples (e.g., one sample per class) labeled. This is the so-called semi-supervised domain adaptation (SSDA) problem and the few labeled target samples are termed as “landmarks”. To explore the full potential of landmarks, we incorporate a prototypical alignment (PA) module which calculates a target prototype for each class from the landmarks; source samples are then aligned with the target prototype from the same class. To further alleviate label scarcity, we propose a data augmentation based solution. Speciﬁcally, we severely perturb the labeled images, making PA non-trivial to achieve and thus promoting model generalizability. Moreover, we apply consistency learning on unlabeled target images, by perturbing each image with light transformations and strong transformations. Then, the strongly perturbed image can enjoy “supervised-like” training using the pseudo label inferred from the lightly perturbed one. Experiments show that the proposed method, though simple, reaches signiﬁcant performance gains over state-of-the-art methods, and enjoys the ﬂexibility of being able to serve as a plug-and-play component to various existing UDA methods and improve adaptation performance with landmarks provided. Our code is available at https://github.com/kailigo/pacl .

LanDA: Language-Guided Multi-Source Domain Adaptation

Transformer-Based Multi-Source Domain Adaptation Without Source Data.

Multi-Source Domain Adaptation through Dataset Dictionary Learning in Wasserstein Space

Local Weight Coupled Network: Multi-Modal Unequal Semi-Supervised Domain Adaptation.

Mdalu: Multi-Source Domain Adaptation and Label Unification with Partial Datasets

Domain Attention Consistency for Multi-Source Domain Adaptation

Learning to Combine: Knowledge Aggregation for Multi-source Domain Adaptation

MultiDAN: Unsupervised, Multistage, Multisource and Multitarget Domain Adaptation for Semantic Segmentation of Remote Sensing Images

Multi-Source Domain Adaptation with Collaborative Learning for Semantic Segmentation

Multi-Source Domain Adaptation meets Dataset Distillation through Dataset Dictionary Learning

Multi-Source Video Domain Adaptation with Temporal Attentive Moment Alignment Network

Multi-source Domain Adaptation in the Deep Learning Era: A Systematic Survey

Iterative Refinement for Multi-Source Visual Domain Adaptation (extended Abstract).

Online Multi-Source Domain Adaptation through Gaussian Mixtures and Dataset Dictionary Learning

Dataset Dictionary Learning in a Wasserstein Space for Federated Domain Adaptation

Learning Distribution-Matched Landmarks For Unsupervised Domain Adaptation

Multi-Source Unsupervised Domain Adaptation Via Pseudo Target Domain

Multi-Target Domain Adaptation with Collaborative Consistency Learning

Semi-Supervised Domain Adaptation with Prototypical Alignment and Consistency Learning

Multi-Source and Multi-Target Domain Adaptation Based on Dynamic Generator With Attention

Empowering Source-Free Domain Adaptation with MLLM-driven Curriculum Learning