SCAN

This repository provides all intermediate files required to reproduce the results of “Endowing Vision-Language Models with System 2 Thinking for Fine-Grained Visual Recognition” (AAAI 2026), including CLIP predictions, category descriptions, image textual descriptions, reranker outputs.

本仓库提供了复现《赋予视觉语言模型系统2思维以实现细粒度视觉识别》（发表于美国人工智能协会2026年年会，AAAI）一文所需的全部中间文件，包括CLIP（Contrastive Language-Image Pre-training）预测结果、类别描述、图像文本描述以及重排序器（reranker）输出结果。