Unified Deep Supervised Domain Adaptation and Generalization

TL;DR

This paper introduces a unified deep learning framework for supervised domain adaptation and generalization, leveraging Siamese networks to efficiently align and separate visual domain features with minimal labeled data.

Contribution

It proposes a novel Siamese-based embedding method that effectively handles limited labeled target data for both domain adaptation and generalization tasks.

Findings

High speed adaptation with very few labeled samples

Effective semantic alignment and separation of domains

Promising experimental results in both tasks

Abstract

This work provides a unified framework for addressing the problem of visual supervised domain adaptation and generalization with deep models. The main idea is to exploit the Siamese architecture to learn an embedding subspace that is discriminative, and where mapped visual domains are semantically aligned and yet maximally separated. The supervised setting becomes attractive especially when only few target data samples need to be labeled. In this scenario, alignment and separation of semantic probability distributions is difficult because of the lack of data. We found that by reverting to point-wise surrogates of distribution distances and similarities provides an effective solution. In addition, the approach has a high speed of adaptation, which requires an extremely low number of labeled target training samples, even one per category can be effective. The approach is extended to domain generalization. For both applications the experiments show very promising results.