Structure-Aware Face Clustering on a Large-Scale Graph with $\bf{10^{7}}$ Nodes

TL;DR

The paper introduces STAR-FC, a structure-aware face clustering method that effectively handles large-scale graphs with over 10 million nodes, achieving high accuracy and efficiency in face annotation tasks.

Contribution

The paper presents a novel large-scale face clustering approach that combines structure-preserved subgraph sampling with efficient full-graph inference, enabling training on 20 million nodes and superior performance.

Findings

Achieved 91.97 F-score on partial MS1M within 310 seconds.

Successfully trained on a 20 million node graph, surpassing previous large-scale training limits.

Provided a strong baseline for large-scale face clustering with high accuracy and efficiency.

Abstract

Face clustering is a promising method for annotating unlabeled face images. Recent supervised approaches have boosted the face clustering accuracy greatly, however their performance is still far from satisfactory. These methods can be roughly divided into global-based and local-based ones. Global-based methods suffer from the limitation of training data scale, while local-based ones are difficult to grasp the whole graph structure information and usually take a long time for inference. Previous approaches fail to tackle these two challenges simultaneously. To address the dilemma of large-scale training and efficient inference, we propose the STructure-AwaRe Face Clustering (STAR-FC) method. Specifically, we design a structure-preserved subgraph sampling strategy to explore the power of large-scale training data, which can increase the training data scale from ${10^{5}}$ to ${10^{7}}$. During inference, the STAR-FC performs efficient full-graph clustering with two steps: graph parsing and graph refinement. And the concept of node intimacy is introduced in the second step to mine the local structural information. The STAR-FC gets 91.97 pairwise F-score on partial MS1M within 310s which surpasses the state-of-the-arts. Furthermore, we are the first to train on very large-scale graph with 20M nodes, and achieve superior inference results on 12M testing data. Overall, as a simple and effective method, the proposed STAR-FC provides a strong baseline for large-scale face clustering. Code is available at \url{https://sstzal.github.io/STAR-FC/}.