search

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

Input image should be normalized as follows: #18

Closed ghost closed 4 years ago

ghost commented 4 years ago

Input image should be normalized as follows:

normalize = transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])

can you tell me how to do it in detail and give a example? thanks

ghost commented 4 years ago

and if i do not normalized, how about the result ?

gasvn commented 4 years ago

You can refer to the example code of training ImageNet of pytorch to see how to implement normalization. https://github.com/pytorch/examples/blob/master/imagenet/main.py

gasvn commented 4 years ago

and if i do not normalized, how about the result ?

If the data distribution is much different from the data used for training the pretrained model, the parameters in pretrained model will have limited help when applied to new data.