Semi-Supervised Learning using Siamese Networks

TL;DR

This paper introduces a semi-supervised learning approach using Siamese networks to learn discriminative embeddings, enabling effective labeling of unlabeled data through iterative self-training and local consistency methods.

Contribution

It proposes a novel semi-supervised training method based on similarity learning with Siamese networks, incorporating iterative self-training and local consistency for improved unlabeled data labeling.

Findings

Effective semi-supervised learning with Siamese networks.

Iterative self-training improves classification accuracy.

Local learning with global consistency enhances unlabeled data labeling.

Abstract

Neural networks have been successfully used as classification models yielding state-of-the-art results when trained on a large number of labeled samples. These models, however, are more difficult to train successfully for semi-supervised problems where small amounts of labeled instances are available along with a large number of unlabeled instances. This work explores a new training method for semi-supervised learning that is based on similarity function learning using a Siamese network to obtain a suitable embedding. The learned representations are discriminative in Euclidean space, and hence can be used for labeling unlabeled instances using a nearest-neighbor classifier. Confident predictions of unlabeled instances are used as true labels for retraining the Siamese network on the expanded training set. This process is applied iteratively. We perform an empirical study of this iterative self-training algorithm. For improving unlabeled predictions, local learning with global consistency [22] is also evaluated.