An image clustering algorithm based on depth semantic embedding

TL;DR: In this article, an image clustering algorithm based on depth semantic embedding is proposed, which divides an image data set into a training set and a test set, and obtaining respective data feature spaces.

Abstract: The invention relates to an image clustering algorithm based on depth semantic embedding, comprising the following steps of 1 dividing an image data set into a training set and a test set, and obtaining respective data feature spaces; 2 obtaining a mapping function of the image data from the data feature space of the training set obtained in the step 1 to the semantic space of the training set, and obtaining the semantic space of the test set through the mapping function; 3 taking the result obtained in the step 2 as an input layer, fusing and reducing dimensions through self-coding to obtaina low-dimensional embedded space with semantic information and original features; 4 clustering in the low-dimensional embedded space with semantic information and original features obtained in the step 4 by using the KL divergence function, and if the KL divergence function converges, ending the clustering; otherwise, returning to step 3 and updating the input layer of Step 3. The algorithm of theinvention effectively improves the discriminability of the data features and improves the clustering effect.