yeungchenwa / FontDiffuser

[AAAI2024] FontDiffuser: One-Shot Font Generation via Denoising Diffusion with Multi-Scale Content Aggregation and Style Contrastive Learning
https://yeungchenwa.github.io/fontdiffuser-homepage/
231 stars 21 forks source link

InfoNCE #50

Open gfdbm opened 1 month ago

gfdbm commented 1 month ago

您好,我想询问一下关于第二阶段的对比损失InfoNCE具体代码是怎么样的,感谢

laiyingxin2 commented 3 weeks ago

same question