Prototypical Partial Optimal Transport for Universal Domain Adaptation

Abstract

Universal domain adaptation (UniDA) aims to transfer knowledge from a labeled source domain to an unlabeled target domain without requiring the same label sets of both domains. The existence of domain and category shift makes the task challenging and requires us to distinguish "known" samples (i.e., samples whose labels exist in both domains) and "unknown" samples (i.e., samples whose labels exist in only one domain) in both domains before reducing the domain gap. In this paper, we consider the problem from the point of view of distribution matching which we only need to align two distributions partially. A novel approach, dubbed mini-batch Prototypical Partial Optimal Transport (m-PPOT), is proposed to conduct partial distribution alignment for UniDA. In training phase, besides minimizing m-PPOT, we also leverage the transport plan of m-PPOT to reweight source prototypes and target samples, and design reweighted entropy loss and reweighted cross-entropy loss to distinguish "known" and "unknown" samples. Experiments on four benchmarks show that our method outperforms the previous state-of-the-art UniDA methods.

References in corpus (12)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Learning Transferable Features with Deep Adaptation Networks
• Unsupervised Domain Adaptation by Backpropagation
• VisDA: The Visual Domain Adaptation Challenge
• Prevalence of Neural Collapse during the terminal phase of deep learning training
• Deep CORAL: Correlation Alignment for Deep Domain Adaptation
• Bridging Theory and Algorithm for Domain Adaptation
• Partial Transfer Learning with Selective Adversarial Networks
• Transferrable Prototypical Networks for Unsupervised Domain Adaptation
• Multilevel Clustering via Wasserstein Means
• Learning to Transfer Examples for Partial Domain Adaptation
• Unbalanced minibatch Optimal Transport; applications to Domain Adaptation