Vision Transformer-based Feature Extraction for Generalized Zero-Shot Learning

TL;DR

This paper introduces a Vision Transformer-based method for generalized zero-shot learning that leverages patch and CLS features with an attribute attention module to improve unseen class recognition.

Contribution

It proposes a novel attribute attention module (AAM) that effectively aggregates attribute-related information from patch features in ViT for GZSL.

Findings

Outperforms state-of-the-art GZSL methods on benchmark datasets.

Effectively preserves local image information through ViT patches.

Demonstrates significant accuracy improvements in recognizing unseen classes.

Abstract

Generalized zero-shot learning (GZSL) is a technique to train a deep learning model to identify unseen classes using the image attribute. In this paper, we put forth a new GZSL approach exploiting Vision Transformer (ViT) to maximize the attribute-related information contained in the image feature. In ViT, the entire image region is processed without the degradation of the image resolution and the local image information is preserved in patch features. To fully enjoy these benefits of ViT, we exploit patch features as well as the CLS feature in extracting the attribute-related image feature. In particular, we propose a novel attention-based module, called attribute attention module (AAM), to aggregate the attribute-related information in patch features. In AAM, the correlation between each patch feature and the synthetic image attribute is used as the importance weight for each patch. From extensive experiments on benchmark datasets, we demonstrate that the proposed technique outperforms the state-of-the-art GZSL approaches by a large margin.