Hardness Sampling for Self-Training Based Transductive Zero-Shot Learning

TL;DR

This paper introduces a hardness sampling approach within a self-training framework for transductive zero-shot learning, improving the utilization of unseen-class samples and outperforming state-of-the-art methods on benchmarks.

Contribution

It proposes a novel hardness sampling method and a flexible self-training framework for T-ZSL, addressing the effective use of unseen data and establishing a new baseline for T-GZSL.

Findings

The proposed methods outperform existing state-of-the-art on three benchmarks.

Hardness sampling improves the diversity and difficulty balance of selected samples.

The framework can embed various ZSL methods and enhances their performance.

Abstract

Transductive zero-shot learning (T-ZSL) which could alleviate the domain shift problem in existing ZSL works, has received much attention recently. However, an open problem in T-ZSL: how to effectively make use of unseen-class samples for training, still remains. Addressing this problem, we first empirically analyze the roles of unseen-class samples with different degrees of hardness in the training process based on the uneven prediction phenomenon found in many ZSL methods, resulting in three observations. Then, we propose two hardness sampling approaches for selecting a subset of diverse and hard samples from a given unseen-class dataset according to these observations. The first one identifies the samples based on the class-level frequency of the model predictions while the second enhances the former by normalizing the class frequency via an approximate class prior estimated by an explored prior estimation algorithm. Finally, we design a new Self-Training framework with Hardness Sampling for T-ZSL, called STHS, where an arbitrary inductive ZSL method could be seamlessly embedded and it is iteratively trained with unseen-class samples selected by the hardness sampling approach. We introduce two typical ZSL methods into the STHS framework and extensive experiments demonstrate that the derived T-ZSL methods outperform many state-of-the-art methods on three public benchmarks. Besides, we note that the unseen-class dataset is separately used for training in some existing transductive generalized ZSL (T-GZSL) methods, which is not strict for a GZSL task. Hence, we suggest a more strict T-GZSL data setting and establish a competitive baseline on this setting by introducing the proposed STHS framework to T-GZSL.