请问用CLIP做多标签图像分类，训练数据怎么处理？

OFA-Sys / Chinese-CLIP

Chinese version of CLIP which achieves Chinese cross-modal retrieval and representation generation.

MIT License

4.46k stars 462 forks source link

Closed dhdhdeng closed 4 months ago

dhdhdeng commented 1 year ago

请问用CLIP做多标签图像分类，训练数据怎么处理？一张图重复很多遍，只对应单个标签，还是把一张图的所有标签拼接在一起？

yangapku commented 1 year ago

建议选择前者哈，或者可以两者结合起来试一试