张量尺寸不匹配问题

Res2Net / Res2Net-PretrainedModels

(ImageNet pretrained models) The official pytorch implemention of the TPAMI paper "Res2Net: A New Multi-scale Backbone Architecture"

https://mmcheng.net/res2net/

1.07k stars 215 forks source link

Closed kirkzZ closed 2 years ago

kirkzZ commented 2 years ago

你好请问 for i in range(self.nums): if i==0 or self.stype=='stage': sp = spx[i] else: sp = sp + spx[i] //这里因为两个特征的尺寸不一样并不能加在一起啊？一个是64一个是128

gasvn commented 2 years ago

是的，如果你使用了stride=2的conv导致特征尺寸不一样，res2net module默认退化成group conv的形式。你也可以考虑用pooling的方式对齐特征尺寸