Abstract:Domain transfer learning aims to learn common data representations from a source domain and a target domain so that the source domain data can help the classification of the target domain. Conventional transfer representation learning imposes the distributions of source and target domain representations to be similar, which heavily relies on the characterization of the distributions of domains and the distribution matching criteria. In this paper, we proposed a novel framework for domain transfer representation learning. Our motive is to make the learned representations of data points independent from the domains which they belong to. In other words, from an optimal cross-domain representation of a data point, it is difficult to tell which domain it is from. In this way, the learned representations can be generalized to different domains. To measure the dependency between the representations and the corresponding domain which the data points belong to, we propose to use the mutual information between the representations and the domain-belonging indicators. By minimizing such mutual information, we learn the representations which are independent from domains. We build a classwise deep convolutional network model as a representation model and maximize the margin of each data point of the corresponding class, which is defined over the intraclass and interclass neighborhood. To learn the parameters of the model, we construct a unified minimization problem where the margins are maximized while the representation-domain mutual information is minimized. In this way, we learn representations which are not only discriminate but also independent from domains. An iterative algorithm based on the Adam optimization method is proposed to solve the minimization to learn the classwise deep model parameters and the cross-domain representations simultaneously. Extensive experiments over benchmark datasets show its effectiveness and advantage over existing domain transfer learning methods.

Self-Supervised Representation Learning From Multi-Domain Data

Joint Domain Alignment and Discriminative Feature Learning for Unsupervised Deep Domain Adaptation

Learning Domain-Independent Deep Representations by Mutual Information Minimization

Learning Representations by Maximizing Mutual Information Across Views

Invariant and Sufficient Supervised Representation Learning

Constrained Multiview Representation for Self-supervised Contrastive Learning

Multi-level Self-supervised Representation Learning Via Triple-way Attention Fusion and Local Similarity Optimization

Depthwise Convolution is All You Need for Learning Multiple Visual Domains

Cross-domain Representation-Learning Framework with Combination of Class-Separate and Domain-Merge Objectives

Research on Cross-domain Representation Learning Based on Multi-network Space Fusion

Multi-Trusted Cross-Modal Information Bottleneck for 3D Self-Supervised Representation Learning

Representation Learning for Imbalanced Cross-Domain Classification

Unsupervised Multi-Domain Image Translation with Domain-Specific Encoders/Decoders

On Learning Invariant Representation for Domain Adaptation

Domain-Aware Self-Attention for Multi-Domain Neural Machine Translation

Self-Supervised Representation Learning by Rotation Feature Decoupling

Unsupervised person re-identification via multi-domain joint learning

Learning Generalizable Visual Representations via Self-Supervised Information Bottleneck.

Representation Learning with Multiple Lipschitz-Constrained Alignments on Partially-Labeled Cross-Domain Data.

Universal Representation Learning from Multiple Domains for Few-shot Classification

Supervised Representation Learning for Multi-Label Classification