Improving Image Clustering through Sample Ranking and Its Application to remote--sensing images

TL;DR

This paper introduces a novel sample ranking method to enhance image clustering, especially for remote sensing images, by weighting samples based on their likelihood of belonging to current clusters, leading to significant accuracy improvements.

Contribution

The paper proposes a new sample ranking and weighting strategy to improve self-supervised image clustering models, with demonstrated effectiveness on remote sensing datasets.

Findings

Achieved accuracy improvements of 2.1% to 15.9% over state-of-the-art models.

Effectively applied the method to various remote sensing datasets.

Enhanced clustering performance by incorporating sample confidence in the training process.

Abstract

Image clustering is a very useful technique that is widely applied to various areas, including remote sensing. Recently, visual representations by self-supervised learning have greatly improved the performance of image clustering. To further improve the well-trained clustering models, this paper proposes a novel method by first ranking samples within each cluster based on the confidence in their belonging to the current cluster and then using the ranking to formulate a weighted cross-entropy loss to train the model. For ranking the samples, we developed a method for computing the likelihood of samples belonging to the current clusters based on whether they are situated in densely populated neighborhoods, while for training the model, we give a strategy for weighting the ranked samples. We present extensive experimental results that demonstrate that the new technique can be used to improve the State-of-the-Art image clustering models, achieving accuracy performance gains ranging from $2.1\%$ to $15.9\%$. Performing our method on a variety of datasets from remote sensing, we show that our method can be effectively applied to remote--sensing images.