Revealing the Proximate Long-Tail Distribution in Compositional Zero-Shot Learning

TL;DR

This paper addresses compositional zero-shot learning by transforming it into a class imbalance problem, leveraging class priors to improve class prototype discrimination and achieve state-of-the-art results.

Contribution

It introduces a novel perspective by modeling CZSL as a proximate class imbalance problem and incorporates class priors into training and inference.

Findings

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

Effectively mitigates visual bias caused by compositional diversity.

No additional parameters required for the proposed method.

Abstract

Compositional Zero-Shot Learning (CZSL) aims to transfer knowledge from seen state-object pairs to novel unseen pairs. In this process, visual bias caused by the diverse interrelationship of state-object combinations blurs their visual features, hindering the learning of distinguishable class prototypes. Prevailing methods concentrate on disentangling states and objects directly from visual features, disregarding potential enhancements that could arise from a data viewpoint. Experimentally, we unveil the results caused by the above problem closely approximate the long-tailed distribution. As a solution, we transform CZSL into a proximate class imbalance problem. We mathematically deduce the role of class prior within the long-tailed distribution in CZSL. Building upon this insight, we incorporate visual bias caused by compositions into the classifier's training and inference by estimating it as a proximate class prior. This enhancement encourages the classifier to acquire more discernible class prototypes for each composition, thereby achieving more balanced predictions. Experimental results demonstrate that our approach elevates the model's performance to the state-of-the-art level, without introducing additional parameters. Our code is available at \url{https://github.com/LanchJL/ProLT-CZSL}.