Instance Similarity Learning for Unsupervised Feature Representation

TL;DR

This paper introduces an unsupervised instance similarity learning method using GANs to better capture semantic relationships in feature space, leading to improved image classification performance.

Contribution

It presents a novel GAN-based approach for mining feature manifolds to learn semantic similarities without supervision, enhancing feature representation quality.

Findings

Outperforms state-of-the-art methods on image classification tasks.

Effectively captures semantic similarity among instances.

Demonstrates the superiority of GAN-based manifold mining.

Abstract

In this paper, we propose an instance similarity learning (ISL) method for unsupervised feature representation. Conventional methods assign close instance pairs in the feature space with high similarity, which usually leads to wrong pairwise relationship for large neighborhoods because the Euclidean distance fails to depict the true semantic similarity on the feature manifold. On the contrary, our method mines the feature manifold in an unsupervised manner, through which the semantic similarity among instances is learned in order to obtain discriminative representations. Specifically, we employ the Generative Adversarial Networks (GAN) to mine the underlying feature manifold, where the generated features are applied as the proxies to progressively explore the feature manifold so that the semantic similarity among instances is acquired as reliable pseudo supervision. Extensive experiments on image classification demonstrate the superiority of our method compared with the state-of-the-art methods. The code is available at https://github.com/ZiweiWangTHU/ISL.git.