-
Notifications
You must be signed in to change notification settings - Fork 1
/
optimize.py
158 lines (135 loc) · 5.88 KB
/
optimize.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
import torch
import torch.optim as optim
import torch.nn as nn
from PIL import Image
from torchvision import transforms
from torchvision.utils import save_image
from math import log10
import os
import re
from typing import List, Optional, Tuple, Union
from glob import glob
import dnnlib
import numpy as np
import legacy
import mapping_net_opt
import pickle
import lpips
def PSNR(mse, flag = 0):
#flag = 0 if a single image is used and 1 if loss for a batch of images is to be calculated
if flag == 0:
psnr = 10 * log10(1 / mse.item())
return psnr
def parse_range(s: Union[str, List]) -> List[int]:
'''Parse a comma separated list of numbers or ranges and return a list of ints.
Example: '1,2,5-10' returns [1, 2, 5, 6, 7]
'''
if isinstance(s, list): return s
ranges = []
range_re = re.compile(r'^(\d+)-(\d+)$')
for p in s.split(','):
m = range_re.match(p)
if m:
ranges.extend(range(int(m.group(1)), int(m.group(2))+1))
else:
ranges.append(int(p))
return ranges
def parse_vec2(s: Union[str, Tuple[float, float]]) -> Tuple[float, float]:
'''Parse a floating point 2-vector of syntax 'a,b'.
Example:
'0,1' returns (0,1)
'''
if isinstance(s, tuple): return s
parts = s.split(',')
if len(parts) == 2:
return (float(parts[0]), float(parts[1]))
raise ValueError(f'cannot parse 2-vector {s}')
def make_transform(translate: Tuple[float,float], angle: float):
m = np.eye(3)
s = np.sin(angle/360.0*np.pi*2)
c = np.cos(angle/360.0*np.pi*2)
m[0][0] = c
m[0][1] = s
m[0][2] = translate[0]
m[1][0] = -s
m[1][1] = c
m[1][2] = translate[1]
return m
#----------------------------------------------------------------------------
def generate_images(network_pkl, z, translate=(0,0), rotate=0):
device = torch.device('cuda')
with dnnlib.util.open_url(network_pkl) as f:
G = legacy.load_network_pkl(f)['G_ema'].to(device) # type: ignore
if hasattr(G.synthesis, 'input'):
m = make_transform(translate, rotate)
m = np.linalg.inv(m)
G.synthesis.input.transform.copy_(torch.from_numpy(m))
img = G.synthesis(z)
return img
def embedding_function():
device = torch.device('cuda')
os.environ["CUDA_VISIBLE_DEVICES"] = "1"
transform = transforms.Compose([transforms.Resize(1024), transforms.ToTensor()])
latents_list = []
imgdir = sorted(glob('./frame_angry_cut/*.png'))
for i in range(954):
print(imgdir[i])
image = Image.open(imgdir[i])
image = image.convert("RGB")
image = transform(image)
image = image.unsqueeze(0)
image = image.to(device)
# MSE loss object
MSE_loss = nn.MSELoss(reduction="mean")
model = mapping_net_opt.MappingNetwork()
model.load_state_dict(torch.load("./model_last_5000.pth"))
model.to(device)
model.eval()
if i == 0 or i == 115 or i == 232 or i == 336 or i == 447 or i == 569 or i == 699 or i == 822: # angry
# if i == 0 or i == 105 or i == 212 or i == 316 or i == 419 or i == 529 or i == 648 or i == 773: # calm
# if i == 0 or i == 115 or i == 231 or i == 347 or i == 461 or i == 589 or i == 711 or i == 831: # disgust
# if i == 0 or i == 109 or i == 217 or i == 319 or i == 429 or i == 578 or i == 704 or i == 822: # fearful
# if i == 0 or i == 103 or i == 206 or i == 311 or i == 421 or i == 530 or i == 638 or i == 755: # happy
# if i == 0 or i == 98 or i == 197 or i == 294: # neutral
# if i == 0 or i == 114 or i == 220 or i == 324 or i == 424 or i == 537 or i == 647 or i == 758: # sad
# if i == 0 or i == 101 or i == 198 or i == 303 or i == 400 or i == 502 or i == 608 or i == 712: #surprise
latents = model.forward(image)
latents_copy = latents.clone().detach()
latents_copy = torch.reshape(latents_copy, (1, 16, 512))
latents_copy.requires_grad = True
#latents_rand = torch.zeros((1, 512), requires_grad=True, device=device)
# Optimizer to change latent code in each backward step
optimizer = optim.Adam({latents_copy}, lr=0.01, betas=(0.9, 0.999), eps=1e-8)
#optimizer = optim.Adam({latents_rand}, lr=0.01, betas=(0.9, 0.999), eps=1e-8)
# Loop to optimise latent vector to match the generated image to input image
loss_ = []
loss_psnr = []
for e in range(1500):
optimizer.zero_grad()
syn_img = generate_images(network_pkl="https://api.ngc.nvidia.com/v2/models/nvidia/research/stylegan3/versions/1/files/stylegan3-r-ffhq-1024x1024.pkl", z=latents_copy)
#syn_img = generate_images(network_pkl="https://api.ngc.nvidia.com/v2/models/nvidia/research/stylegan3/versions/1/files/stylegan3-r-ffhq-1024x1024.pkl", z=latents_rand)
syn_img = (syn_img + 1.0) / 2.0
mse = MSE_loss(syn_img, image)
loss_fn_alex = lpips.LPIPS(net='alex')
loss_fn_alex.to(device)
p = loss_fn_alex(syn_img, image)
psnr = PSNR(mse, flag=0)
loss = mse + p
if loss < 0.03:
save_image(syn_img.clamp(0, 1), "./opt_angry/optimized_angry{frame:04d}_{epoch}.png".format(frame=i, epoch=e + 1))
break
loss.backward()
optimizer.step()
loss_np = loss.detach().cpu().numpy()
loss_np = loss_np[0,0,0,0]
print(loss_np)
loss_psnr.append(psnr)
loss_.append(loss_np)
if e+1 == 1500:
save_image(syn_img.clamp(0, 1), "./opt_angry/optimized_angry{frame:04d}_{epoch}.png".format(frame=i, epoch=e + 1))
latents_copy_list = latents_copy.detach().clone()
latents_list.append(latents_copy_list)
with open('latents_angry.pkl', 'wb') as f:
pickle.dump(latents_list, f)
if __name__ == "__main__":
embedding_function()