NVlabs / DG-Net

:couple: Joint Discriminative and Generative Learning for Person Re-identification. CVPR'19 (Oral) :couple:
https://www.zdzheng.xyz/publication/Joint-di2019
Other
1.27k stars 230 forks source link

如何采用训练好的模型去生成图片 #65

Closed SeibertronSS closed 2 years ago

SeibertronSS commented 3 years ago

您好,我和我的同事对您的工作十分感兴趣。现在我们遇到了一个问题,我们想要采用您训练好的模型去生成更多的图像,但是我们无法读入您训练好的模型参数。事实上,我们在构建好DG-NET之后,无法读取您在网盘中存入的net_last.pth文件,模型与参数是不匹配的。请问您知道是什么原因吗?

layumi commented 3 years ago

@SeibertronSS 您能提供具体的报错 log么?