Semi-Supervised Hierarchical Open-Set Classification

TL;DR

This paper introduces a semi-supervised hierarchical open-set classification method using a teacher-student framework with subtree pseudo-labels and age-gating, significantly improving performance on large-scale uncurated datasets.

Contribution

It presents a novel semi-supervised approach for hierarchical open-set classification that effectively leverages uncurated data with pseudo-labeling and new mechanisms to handle unknown classes.

Findings

Outperforms self-supervised pretraining and supervised adaptation.

Matches fully supervised performance with only 20 labeled samples per class.

Effective on large-scale uncurated datasets like iNaturalist19.

Abstract

Hierarchical open-set classification handles previously unseen classes by assigning them to the most appropriate high-level category in a class taxonomy. We extend this paradigm to the semi-supervised setting, enabling the use of large-scale, uncurated datasets containing a mixture of known and unknown classes to improve the hierarchical open-set performance. To this end, we propose a teacher-student framework based on pseudo-labeling. Two key components are introduced: 1) subtree pseudo-labels, which provide reliable supervision in the presence of unknown data, and 2) age-gating, a mechanism that mitigates overconfidence in pseudo-labels. Experiments show that our framework outperforms self-supervised pretraining followed by supervised adaptation, and even matches the fully supervised counterpart when using only 20 labeled samples per class on the iNaturalist19 benchmark. Our code is available at https://github.com/walline/semihoc.