Abstract:Heterogeneous domain adaptation (HDA) aims to address the transfer learning problems where the source domain and target domain are represented by heterogeneous features. The existing HDA methods based on matrix factorization have been proven to learn transferable features effectively. However, these methods only preserve the original neighbor structure of samples in each domain and do not use the label information to explore the similarity and separability between samples. This would not eliminate the cross-domain bias of samples and may mix cross-domain samples of different classes in the common subspace, misleading the discriminative feature learning of target samples. To tackle the aforementioned problems, we propose a novel matrix factorization-based HDA method called HDA with generalized similarity and dissimilarity regularization (HGSDR). Specifically, we propose a similarity regularizer by establishing the cross-domain Laplacian graph with label information to explore the similarity between cross-domain samples from the identical class. And we propose a dissimilarity regularizer based on the inner product strategy to expand the separability of cross-domain labeled samples from different classes. For unlabeled target samples, we keep their neighbor relationship to preserve the similarity and separability between them in the original space. Hence, the generalized similarity and dissimilarity regularization is built by integrating the above regularizers to facilitate cross-domain samples to form discriminative class distributions. HGSDR can more efficiently match the distributions of the two domains both from the global and sample viewpoints, thereby learning discriminative features for target samples. Extensive experiments on the benchmark datasets demonstrate the superiority of the proposed method against several state-of-the-art methods.

Robust multiple subspaces transfer for heterogeneous domain adaptation

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

Heterogeneous Domain Adaptation With Generalized Similarity and Dissimilarity Regularization

Learning with Augmented Features for Heterogeneous Domain Adaptation

Few-shot Hybrid Domain Adaptation of Image Generator

Domain Transfer via Multiple Sources Regularization.

LSDT: Latent Sparse Domain Transfer Learning for Visual Adaptation

Heterogeneous Multidomain Recommender System Through Adversarial Learning

Semi-supervised Domain Adaptation with Subspace Learning for Visual Recognition

Heterogeneous Domain Adaptation: An Unsupervised Approach

Semi-Supervised Optimal Transport for Heterogeneous Domain Adaptation

Transfer subspace learning for cross-dataset facial expression recognition

Learning a Deep Structural Subspace Across Hyperspectral Scenes with Cross-Domain VAE

Semantic Correlation Transfer for Heterogeneous Domain Adaptation

Manifold transfer subspace learning based on double relaxed discriminative regression

Heterogeneous domain adaptation by semantic distribution alignment network

Domain-Agnostic Prior for Transfer Semantic Segmentation.

Unsupervised Domain Adaptation Using Robust Class-Wise Matching

Multiple Graphs and Low-Rank Embedding for Multi-Source Heterogeneous Domain Adaptation

Fine-grained Representation Alignment for Zero-shot Domain Adaptation

Learning explicitly transferable representations for domain adaptation