Abstract:Partial domain adaptation (PDA) attempts to learn transferable models from a large-scale labeled source domain to a small unlabeled target domain with fewer classes, which has attracted a recent surge of interest in transfer learning. Most conventional PDA approaches endeavor to design delicate source weighting schemes by leveraging target predictions to align cross-domain distributions in the shared class space. Accordingly, two crucial issues are overlooked in these methods. First, target prediction is a double-edged sword, and inaccurate predictions will result in negative transfer inevitably. Second, not all target samples have equal transferability during the adaptation; thus, "ambiguous" target data predicted with high uncertainty should be paid more attentions. In this article, we propose a critical classes and samples discovering network (CSDN) to identify the most relevant source classes and critical target samples, such that more precise cross-domain alignment in the shared label space could be enforced by co-training two diverse classifiers. Specifically, during the training process, CSDN introduces an adaptive source class weighting scheme to select the most relevant classes dynamically. Meanwhile, based on the designed target ambiguous score, CSDN emphasizes more on ambiguous target samples with larger inconsistent predictions to enable fine-grained alignment. Taking a step further, the weighting schemes in CSDN can be easily coupled with other PDA and DA methods to further boost their performance, thereby demonstrating its flexibility. Extensive experiments verify that CSDN attains excellent results compared to state of the arts on four highly competitive benchmark datasets.

Class-Aware Sample Reweighting Optimal Transport for Multi-source Domain Adaptation

HOMDA: High-Order Moment-Based Domain Alignment for Unsupervised Domain Adaptation

Target-Discriminability-Induced Multi-Source-Free Domain Adaptation

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

Multiscale Attention-Based Subdomain Dynamic Adaptation for Cross-Domain Scene Classification

Reliable Weighted Optimal Transport For Unsupervised Domain Adaptation

Revisiting the Domain Shift and Sample Uncertainty in Multi-source Active Domain Transfer

Exploring High-Correlation Source Domain Information for Multi-Source Domain Adaptation in Semantic Segmentation

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

Lighter, Better, Faster Multi-Source Domain Adaptation with Gaussian Mixture Models and Optimal Transport

Attention-based Class-Conditioned Alignment for Multi-Source Domain Adaptation of Object Detectors

Critical Classes and Samples Discovering for Partial Domain Adaptation

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

Multi-Source and Test-Time Domain Adaptation on Multivariate Signals using Spatio-Temporal Monge Alignment

Decomposed-distance weighted optimal transport for unsupervised domain adaptation

Collaborative Multi-source Domain Adaptation Through Optimal Transport

Partial Feature Selection and Alignment for Multi-Source Domain Adaptation

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

Unified Optimal Transport Framework for Universal Domain Adaptation

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

Multi-source Distilling Domain Adaptation