视觉语言表示学习综述

Visual language multimodal representation learning has become a recent research hotspot, as fusing different feature information from the two modalities can greatly improve the performance of models in downstream tasks. Visual language representation learning models have achieved good results in many downstream tasks. This article will first introduce the downstream tasks of visual language representation learning, and then introduce the two methods of visual language representation learning: similarity-based visual language representation learning and pre-trained model-based visual language representation learning. Finally, the development trends of visual language representation learning will be discussed.