Learning to Cluster Faces on an Affinity Graph

TL;DR

This paper introduces a graph convolutional network-based framework for large-scale face clustering, significantly improving clustering accuracy and enhancing face recognition performance by leveraging unlabeled data.

Contribution

The work presents a novel learning-based clustering method using graph neural networks, addressing challenges of large-scale face clustering with complex pattern variations.

Findings

Achieves more accurate face clustering on large datasets

Leads to improved face recognition performance

Demonstrates effectiveness of learning-based clustering over traditional methods

Abstract

Face recognition sees remarkable progress in recent years, and its performance has reached a very high level. Taking it to a next level requires substantially larger data, which would involve prohibitive annotation cost. Hence, exploiting unlabeled data becomes an appealing alternative. Recent works have shown that clustering unlabeled faces is a promising approach, often leading to notable performance gains. Yet, how to effectively cluster, especially on a large-scale (i.e. million-level or above) dataset, remains an open question. A key challenge lies in the complex variations of cluster patterns, which make it difficult for conventional clustering methods to meet the needed accuracy. This work explores a novel approach, namely, learning to cluster instead of relying on hand-crafted criteria. Specifically, we propose a framework based on graph convolutional network, which combines a detection and a segmentation module to pinpoint face clusters. Experiments show that our method yields significantly more accurate face clusters, which, as a result, also lead to further performance gain in face recognition.