qihoo360/LIT-CN
收藏官方服务：
资源简介：
FG-CLIP 2 是一个在英文和中文两种语言中都能够进行细粒度视觉-语言理解的模型。在29个数据集和8个不同的任务上，它始终优于最近的强基线如SigLIP 2和MetaCLIP 2，在两种语言中都取得了迄今为止最佳的报告性能。
FG-CLIP 2 is a bilingual fine-grained vision-language alignment model that demonstrates strong performance across various tasks and datasets, consistently surpassing recent strong baselines such as SigLIP 2 and MetaCLIP 2, achieving the best reported performance to date in both English and Chinese.
提供机构：
qihoo360
