Dual Adversarial Auto-Encoders for Clustering

TL;DR

This paper introduces Dual Adversarial Auto-Encoders, a novel clustering method that enhances unsupervised learning by maximizing likelihood and mutual information, leading to superior performance on complex data structures.

Contribution

The paper proposes Dual-AAE, which combines variational inference and a new regularization to improve clustering accuracy and disentangle image features without supervision.

Findings

Outperforms state-of-the-art clustering methods on four benchmarks.

Achieves clustering accuracy comparable to supervised CNNs with reject option.

Effectively disentangles style and content in images without supervision.

Abstract

As a powerful approach for exploratory data analysis, unsupervised clustering is a fundamental task in computer vision and pattern recognition. Many clustering algorithms have been developed, but most of them perform unsatisfactorily on the data with complex structures. Recently, Adversarial Auto-Encoder (AAE) shows effectiveness on tackling such data by combining Auto-Encoder (AE) and adversarial training, but it cannot effectively extract classification information from the unlabeled data. In this work, we propose Dual Adversarial Auto-encoder (Dual-AAE) which simultaneously maximizes the likelihood function and mutual information between observed examples and a subset of latent variables. By performing variational inference on the objective function of Dual-AAE, we derive a new reconstruction loss which can be optimized by training a pair of Auto-encoders. Moreover, to avoid mode collapse, we introduce the clustering regularization term for the category variable. Experiments on four benchmarks show that Dual-AAE achieves superior performance over state-of-the-art clustering methods. Besides, by adding a reject option, the clustering accuracy of Dual-AAE can reach that of supervised CNN algorithms. Dual-AAE can also be used for disentangling style and content of images without using supervised information.