Unsupervised Cross-Domain Recognition by Identifying Compact Joint Subspaces

TL;DR

This paper presents a flexible unsupervised cross-domain recognition method that identifies compact joint subspaces for each class, improving recognition accuracy across different domains in vision and NLP tasks.

Contribution

The proposed method uniquely captures class-specific variations by constructing joint subspaces, avoiding reliance on global domain shifts, and effectively handling cross-domain recognition.

Findings

Outperforms existing methods on object recognition datasets.

Achieves higher accuracy in sentiment classification tasks.

Demonstrates robustness across vision and NLP domains.

Abstract

This paper introduces a new method to solve the cross-domain recognition problem. Different from the traditional domain adaption methods which rely on a global domain shift for all classes between source and target domain, the proposed method is more flexible to capture individual class variations across domains. By adopting a natural and widely used assumption -- "the data samples from the same class should lay on a low-dimensional subspace, even if they come from different domains", the proposed method circumvents the limitation of the global domain shift, and solves the cross-domain recognition by finding the compact joint subspaces of source and target domain. Specifically, given labeled samples in source domain, we construct subspaces for each of the classes. Then we construct subspaces in the target domain, called anchor subspaces, by collecting unlabeled samples that are close to each other and highly likely all fall into the same class. The corresponding class label is then assigned by minimizing a cost function which reflects the overlap and topological structure consistency between subspaces across source and target domains, and within anchor subspaces, respectively.We further combine the anchor subspaces to corresponding source subspaces to construct the compact joint subspaces. Subsequently, one-vs-rest SVM classifiers are trained in the compact joint subspaces and applied to unlabeled data in the target domain. We evaluate the proposed method on two widely used datasets: object recognition dataset for computer vision tasks, and sentiment classification dataset for natural language processing tasks. Comparison results demonstrate that the proposed method outperforms the comparison methods on both datasets.