When Unsupervised Domain Adaptation Meets Tensor Representations

TL;DR

This paper introduces a novel tensor-based domain adaptation method that directly aligns tensor representations across domains, preserving structure and improving cross-domain visual recognition performance.

Contribution

It proposes a joint optimization framework for aligning tensor representations without vectorization, enhancing domain adaptation especially for deep network features.

Findings

Outperforms state-of-the-art methods in cross-domain visual recognition

Effective for small sample sizes and one-shot domain adaptation

Resists label noise and preserves discriminative features

Abstract

Domain adaption (DA) allows machine learning methods trained on data sampled from one distribution to be applied to data sampled from another. It is thus of great practical importance to the application of such methods. Despite the fact that tensor representations are widely used in Computer Vision to capture multi-linear relationships that affect the data, most existing DA methods are applicable to vectors only. This renders them incapable of reflecting and preserving important structure in many problems. We thus propose here a learning-based method to adapt the source and target tensor representations directly, without vectorization. In particular, a set of alignment matrices is introduced to align the tensor representations from both domains into the invariant tensor subspace. These alignment matrices and the tensor subspace are modeled as a joint optimization problem and can be learned adaptively from the data using the proposed alternative minimization scheme. Extensive experiments show that our approach is capable of preserving the discriminative power of the source domain, of resisting the effects of label noise, and works effectively for small sample sizes, and even one-shot DA. We show that our method outperforms the state-of-the-art on the task of cross-domain visual recognition in both efficacy and efficiency, and particularly that it outperforms all comparators when applied to DA of the convolutional activations of deep convolutional networks.