透过镜中世界：奇异图像的常识一致性评估

Through the Looking Glass: Common Sense Consistency Evaluation of Weird Images

May 12, 2025

作者: Elisei Rykov, Kseniia Petrushina, Kseniia Titova, Anton Razzhigaev, Alexander Panchenko, Vasily Konovalov

cs.AI

摘要

在人工智能研究中，衡量图像的真实性是一项复杂任务。例如，一张描绘男孩在沙漠中使用吸尘器的图片就违背了常识。我们提出了一种名为“镜中窥真”（Through the Looking Glass, TLG）的新方法，利用大型视觉-语言模型（LVLMs）和基于Transformer的编码器来评估图像的常识一致性。通过LVLMs从图像中提取原子事实，我们获得了一系列准确的事实集合。随后，我们在编码后的原子事实上微调了一个紧凑的注意力池化分类器。我们的TLG方法在WHOOPS!和WEIRD数据集上实现了新的最先进性能，同时仅依赖于一个紧凑的微调组件。

English

Measuring how real images look is a complex task in artificial intelligence research. For example, an image of a boy with a vacuum cleaner in a desert violates common sense. We introduce a novel method, which we call Through the Looking Glass (TLG), to assess image common sense consistency using Large Vision-Language Models (LVLMs) and Transformer-based encoder. By leveraging LVLMs to extract atomic facts from these images, we obtain a mix of accurate facts. We proceed by fine-tuning a compact attention-pooling classifier over encoded atomic facts. Our TLG has achieved a new state-of-the-art performance on the WHOOPS! and WEIRD datasets while leveraging a compact fine-tuning component.