Giters

xwjabc / hed

A PyTorch reimplementation of Holistically-Nested Edge Detection

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Problem

long123524 opened this issue · comments

commented

The functions of load_pretrained_caffe and load_vgg16_caffe have some difference? I don't know which one I should use？ Can you give me some suggests?

commented

load_pretrained_caffe() will load all parameters from pretrained HED caffe checkpoint, while load_vgg16_caffe() only loads the weights for VGG network.

If you want to evaluate the pretrained HED caffe checkpoint (e.g. hed_pretrained_bsds.py36pickle), please use load_pretrained_caffe().

If you want to train a new HED model, please use load_vgg16_caffe() to load VGG weights from 5stage-vgg.py36pickle and train the model on BSDS dataset.

See https://github.com/xwjabc/hed/blob/master/utils.py#L90 for more details.