A review on multimodal zero-shot learning

Multimodal learning provides a path to fully utilize all types of information related to the modeling target to provide the model with a global vision. Zero-shot learning (ZSL) is a general solution for incorporating prior knowledge into data-driven models and achieving accurate class identification. The combination of the two, known as multimodal ZSL (MZSL), can fully exploit the advantages of both technologies and is expected to produce models with greater generalization ability. However, the MZSL algorithms and applications have not yet been thoroughly investigated and summarized. This study fills this gap by providing an objective overview of MZSL's definition, typical algorithms, representative applications, and critical issues. This article will not only provide researchers in this field with a comprehensive perspective, but it will also highlight several promising research directions.This article is categorized under:Algorithmic Development > MultimediaTechnologies > ClassificationTechnologies > Machine Learning

Automated summary

Multimodal learning allows for the utilization of various types of information to provide a comprehensive view for modeling targets. Zero-shot learning integrates prior knowledge into data-driven models for accurate class identification. Combining these two approaches, known as multimodal zero-shot learning, can harness the advantages of both and lead to models with better generalization abilities. However, comprehensive research and summaries on multimodal zero-shot learning algorithms and applications are currently lacking. This study aims to bridge this gap by offering an objective overview of the definition, typical algorithms, representative applications, and crucial issues surrounding multimodal zero-shot learning. This article not only provides researchers in this field with a comprehensive perspective but also highlights several promising research directions.