Deep Image Clustering with Category-Style Representation
This work addresses the challenge of improving clustering accuracy in image analysis, which is incremental as it builds on existing deep clustering methods.
The paper tackles the problem of deep image clustering by learning a category-style latent representation that disentangles category information from image style, enabling direct cluster assignment. The proposed approach significantly outperforms state-of-the-art methods on five public datasets.
Deep clustering which adopts deep neural networks to obtain optimal representations for clustering has been widely studied recently. In this paper, we propose a novel deep image clustering framework to learn a category-style latent representation in which the category information is disentangled from image style and can be directly used as the cluster assignment. To achieve this goal, mutual information maximization is applied to embed relevant information in the latent representation. Moreover, augmentation-invariant loss is employed to disentangle the representation into category part and style part. Last but not least, a prior distribution is imposed on the latent representation to ensure the elements of the category vector can be used as the probabilities over clusters. Comprehensive experiments demonstrate that the proposed approach outperforms state-of-the-art methods significantly on five public datasets.