Knowledge-aware Zero-Shot Learning: Survey and Perspective

TL;DR

This paper reviews zero-shot learning techniques that leverage external knowledge, categorizing and comparing different methods, and discusses the potential of symbolic knowledge to improve ZSL and other machine learning tasks.

Contribution

It provides a comprehensive survey of external knowledge-based ZSL methods and offers perspectives on the role of symbolic knowledge in addressing data scarcity.

Findings

Categorization of external knowledge types in ZSL

Comparison of different ZSL methods using external knowledge

Discussion on the future role of symbolic knowledge in ZSL

Abstract

Zero-shot learning (ZSL) which aims at predicting classes that have never appeared during the training using external knowledge (a.k.a. side information) has been widely investigated. In this paper we present a literature review towards ZSL in the perspective of external knowledge, where we categorize the external knowledge, review their methods and compare different external knowledge. With the literature review, we further discuss and outlook the role of symbolic knowledge in addressing ZSL and other machine learning sample shortage issues.