Deep Online Probability Aggregation Clustering

TL;DR

This paper introduces Probability Aggregation Clustering (PAC), a novel online deep clustering method that improves stability and performance by avoiding cluster centers and optimizing probability distributions, demonstrated through extensive experiments.

Contribution

The paper proposes PAC, a centerless clustering algorithm for online deep clustering, and integrates it into a deep visual clustering framework called DPAC, outperforming existing methods.

Findings

PAC exhibits superior robustness and performance in clustering tasks.

DPAC significantly outperforms state-of-the-art deep clustering methods.

The approach enables stable, flexible clustering over mini-batch data.

Abstract

Combining machine clustering with deep models has shown remarkable superiority in deep clustering. It modifies the data processing pipeline into two alternating phases: feature clustering and model training. However, such alternating schedule may lead to instability and computational burden issues. We propose a centerless clustering algorithm called Probability Aggregation Clustering (PAC) to proactively adapt deep learning technologies, enabling easy deployment in online deep clustering. PAC circumvents the cluster center and aligns the probability space and distribution space by formulating clustering as an optimization problem with a novel objective function. Based on the computation mechanism of the PAC, we propose a general online probability aggregation module to perform stable and flexible feature clustering over mini-batch data and further construct a deep visual clustering framework deep PAC (DPAC). Extensive experiments demonstrate that PAC has superior clustering robustness and performance and DPAC remarkably outperforms the state-of-the-art deep clustering methods.