Attribute-Aware Representation Rectification for Generalized Zero-Shot Learning

TL;DR

This paper introduces AR^2, a framework that adaptively rectifies feature extractors in GZSL by leveraging attribute-aware distillation and attribute-guided learning to improve recognition of unseen classes.

Contribution

The proposed AR^2 framework is a novel approach that effectively balances learning new features and preserving valuable features in GZSL, addressing domain bias without catastrophic forgetting.

Findings

AR^2 outperforms existing methods on benchmark datasets.

Unseen-Aware Distillation improves unseen class feature retention.

Attribute-Guided Learning enhances discriminative feature focus.

Abstract

Generalized Zero-shot Learning (GZSL) has yielded remarkable performance by designing a series of unbiased visual-semantics mappings, wherein, the precision relies heavily on the completeness of extracted visual features from both seen and unseen classes. However, as a common practice in GZSL, the pre-trained feature extractor may easily exhibit difficulty in capturing domain-specific traits of the downstream tasks/datasets to provide fine-grained discriminative features, i.e., domain bias, which hinders the overall recognition performance, especially for unseen classes. Recent studies partially address this issue by fine-tuning feature extractors, while may inevitably incur catastrophic forgetting and overfitting issues. In this paper, we propose a simple yet effective Attribute-Aware Representation Rectification framework for GZSL, dubbed $\mathbf{(AR)^{2}}$, to adaptively rectify the feature extractor to learn novel features while keeping original valuable features. Specifically, our method consists of two key components, i.e., Unseen-Aware Distillation (UAD) and Attribute-Guided Learning (AGL). During training, UAD exploits the prior knowledge of attribute texts that are shared by both seen/unseen classes with attention mechanisms to detect and maintain unseen class-sensitive visual features in a targeted manner, and meanwhile, AGL aims to steer the model to focus on valuable features and suppress them to fit noisy elements in the seen classes by attribute-guided representation learning. Extensive experiments on various benchmark datasets demonstrate the effectiveness of our method.