English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

In this paper, we propose a novel, effective and simpler end-to-end imageclustering auto-encoder algorithm: ICAE. The algorithm uses PEDCC (PredefinedEvenly-Distributed Class Centroids) as the clustering centers, which ensuresthe inter-class distance of latent features is maximal, and adds datadistribution constraint, data augmentation constraint, auto-encoderreconstruction constraint and Sobel smooth constraint to improve the clusteringperformance. Specifically, we perform one-to-one data augmentation to learn themore effective features. The data and the augmented data are simultaneouslyinput into the autoencoder to obtain latent features and the augmented latentfeatures whose similarity are constrained by an augmentation loss. Then, makinguse of the maximum mean discrepancy distance (MMD), we combine the latentfeatures and augmented latent features to make their distribution close to thePEDCC distribution (uniform distribution between classes, Dirac distributionwithin the class) to further learn clustering-oriented features. At the sametime, the MSE of the original input image and reconstructed image is used asreconstruction constraint, and the Sobel smooth loss to build generalizationconstraint to improve the generalization ability. Finally, extensiveexperiments on three common datasets MNIST, Fashion-MNIST, COIL20 areconducted. The experimental results show that the algorithm has achieved thebest clustering results so far. In addition, we can use the predefined PEDCCclass centers, and the decoder to clearly generate the samples of each class.The code can be downloaded at https://github.com/zyWang-Power/Clustering!

An Image Clustering Auto-Encoder Based on Predefined Evenly-Distributed Class Centroids and MMD Distance