HUANGLIZI / LViT

[IEEE Transactions on Medical Imaging/TMI] This repo is the official implementation of "LViT: Language meets Vision Transformer in Medical Image Segmentation"
MIT License
283 stars 26 forks source link

text标注生成 #48

Closed xskdjs444 closed 1 month ago

xskdjs444 commented 1 month ago

你好,大佬,感谢您很有启发性的工作。我想问一下如果我想用其他的2D数据集,比如GLAS,BUSI,ISIC这些,如何用mask来生成text标注,可以提供一下对应的生成代码或者思路吗?期待您的回复!

HUANGLIZI commented 1 month ago

Please take a look at the paper, whose text annotation has three parts of information.