mirror of https://github.com/menyifang/DCT-Net
You cannot select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
41 lines
1.3 KiB
Python
41 lines
1.3 KiB
Python
2 years ago
|
import torch
|
||
|
|
from torch import nn
|
||
|
|
from .model_irse import Backbone
|
||
|
|
|
||
|
|
|
||
|
|
class IDLoss(nn.Module):
|
||
|
|
def __init__(self):
|
||
|
|
super(IDLoss, self).__init__()
|
||
|
|
print('Loading ResNet ArcFace')
|
||
|
|
model_paths = '/data/vdb/qingyao/cartoon/mycode/pretrained_models/model_ir_se50.pth'
|
||
|
|
self.facenet = Backbone(input_size=112, num_layers=50, drop_ratio=0.6, mode='ir_se')
|
||
|
|
self.facenet.load_state_dict(torch.load(model_paths))
|
||
|
|
self.face_pool = torch.nn.AdaptiveAvgPool2d((112, 112))
|
||
|
|
self.facenet.eval()
|
||
|
|
|
||
|
|
def extract_feats(self, x):
|
||
|
|
x = x[:, :, 35:223, 32:220] # Crop interesting region
|
||
|
|
x = self.face_pool(x)
|
||
|
|
x_feats = self.facenet(x)
|
||
|
|
return x_feats
|
||
|
|
|
||
|
|
def forward(self, y_hat, x):
|
||
|
|
n_samples = x.shape[0]
|
||
|
|
x_feats = self.extract_feats(x)
|
||
|
|
y_hat_feats = self.extract_feats(y_hat)
|
||
|
|
loss = 0
|
||
|
|
sim_improvement = 0
|
||
|
|
id_logs = []
|
||
|
|
count = 0
|
||
|
|
for i in range(n_samples):
|
||
|
|
diff_input = y_hat_feats[i].dot(x_feats[i])
|
||
|
|
id_logs.append({
|
||
|
|
'diff_input': float(diff_input)
|
||
|
|
})
|
||
|
|
# loss += 1 - diff_target
|
||
|
|
# modify
|
||
|
|
loss += 1 - diff_input
|
||
|
|
count += 1
|
||
|
|
|
||
|
|
return loss / count
|