Semi-supervised deep learning by metric embedding

TL;DR

This paper introduces a semi-supervised deep learning approach using metric embedding to improve classification with limited labeled data by learning discriminative representations suitable for nearest-neighbor classification.

Contribution

It proposes a novel training objective based on deep metric embedding that effectively leverages unlabeled data in semi-supervised learning.

Findings

Improved classification accuracy with limited labeled data.

Effective discriminative embeddings for semi-supervised tasks.

Compatibility with nearest-neighbor classifiers.

Abstract

Deep networks are successfully used as classification models yielding state-of-the-art results when trained on a large number of labeled samples. These models, however, are usually much less suited for semi-supervised problems because of their tendency to overfit easily when trained on small amounts of data. In this work we will explore a new training objective that is targeting a semi-supervised regime with only a small subset of labeled data. This criterion is based on a deep metric embedding over distance relations within the set of labeled samples, together with constraints over the embeddings of the unlabeled set. The final learned representations are discriminative in euclidean space, and hence can be used with subsequent nearest-neighbor classification using the labeled samples.