一种神经分而治之的推理框架，用于从语言复杂的文本中检索图像（英文版）

英文标题：A Neural Divide-and-Conquer Reasoning Framework for Image Retrieval from Linguistically Complex Text中文摘要：本文提出了一种名为 NDCR 的端到端的神经分治推理框架，将语言复杂的文本视为由多个简单命题句组成的复合命题文本，并包含三个主要组件：命题生成器、基于预训练 VLM 的视觉语言交互器以及神经符号推理器，该框架在复杂的图像 - 文本推理问题中显著提高了性能。英文摘要：Pretrained Vision-Language Models (VLMs) have achieved remarkable performancein image retrieval from text. However, their performance drops drastically whenconfronted with linguistically complex texts that they struggle to comprehend.Inspired by the Divide