Consistent Supervised-Unsupervised Alignment for Generalized Category Discovery

TL;DR

This paper introduces a novel framework for Generalized Category Discovery that uses fixed geometric prototypes and alignment losses to improve the separation and discovery of both known and novel categories.

Contribution

The proposed NC-GCD framework employs fixed ETF prototypes and a consistency matcher to unify optimization objectives and enhance category separation in GCD tasks.

Findings

Achieves state-of-the-art performance on GCD benchmarks.

Significantly improves accuracy on novel categories.

Ensures stable label assignments across clustering iterations.

Abstract

Generalized Category Discovery (GCD) focuses on classifying known categories while simultaneously discovering novel categories from unlabeled data. However, previous GCD methods face challenges due to inconsistent optimization objectives and category confusion. This leads to feature overlap and ultimately hinders performance on novel categories. To address these issues, we propose the Neural Collapse-inspired Generalized Category Discovery (NC-GCD) framework. By pre-assigning and fixing Equiangular Tight Frame (ETF) prototypes, our method ensures an optimal geometric structure and a consistent optimization objective for both known and novel categories. We introduce a Consistent ETF Alignment Loss that unifies supervised and unsupervised ETF alignment and enhances category separability. Additionally, a Semantic Consistency Matcher (SCM) is designed to maintain stable and consistent label assignments across clustering iterations. Our method achieves strong performance on multiple GCD benchmarks, significantly enhancing novel category accuracy and demonstrating its effectiveness.