-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathipref_main.py
179 lines (155 loc) · 10.3 KB
/
ipref_main.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
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
import os
import csv
import argparse
import random
import json
from diffusers.utils import load_image
import torch
import torch.nn.functional as F
import numpy as np
import megfile
import lpips
from diffsim.diffsim import DiffSim, process_image
from diffsim.diffsim_xl import diffsim_xl
from diffsim.diffsim_dit import diffsim_DiT
from metrics.clip_i import CLIPScore
from metrics.dino import Dinov2Score, DinoScore
from metrics.foreground_feature_averaging import ForegroundFeatureAveraging
from argprocess import arg_parse
def ipref1(args, image_path, original_path, device): # Same reference image, different IP weight
random.seed(args.seed)
# init sd model
prompt = args.prompt
if args.metric == 'diffsim' or args.metric == 'diffeats' or args.metric == 'ensemble':
diffsim = DiffSim(torch.float16, device, args.ip_adapter)
if args.metric == 'diffsim_xl':
diffsim_xl_score = diffsim_xl(torch.float16, device, args.ip_adapter)
if args.metric == 'dit':
diffsim_dit = diffsim_DiT(args.image_size, args.target_step, device)
if 'clip' in args.metric or args.metric == 'ensemble':
clip_score = CLIPScore(device=device)
if args.metric == 'dino' or args.metric == 'dino_cross' or args.metric == 'dinofeats' or args.metric == 'ensemble':
dino_score = Dinov2Score(device=device)
if args.metric == 'dinov1':
dino_score = DinoScore(device=device)
if 'cute' in args.metric:
cute_score = ForegroundFeatureAveraging(device=device)
if 'lpips' in args.metric:
lpips_score = lpips.LPIPS(net='vgg')
with torch.no_grad():
total_samples = 0
correct_predictions = 0
prompt = "High quality image"
# step = args.target_step
print(f"=========seed {args.seed}=========")
print(f"Experiment on {args.target_block}, layer {args.target_layer}, timestep {args.target_step}, image size {args.image_size}:")
# load the annotation file
for cls in os.listdir(args.image_path): # iterate through all the IP classes
cls_dir = os.path.join(args.image_path, cls)
compare_pairs = [("1.0.png", "0.6.png"), ("0.8.png", "0.4.png"), ("0.6.png", "0.3.png"), ("0.4.png", "0.35.png"), ("0.3.png", "0.2.png")]
ref_path = os.path.join(original_path, f"{cls}.JPG")
for (image1, image2) in compare_pairs:
image1_path = os.path.join(cls_dir, image1)
image2_path = os.path.join(cls_dir, image2)
if args.metric == 'diffsim':
diff_ab = diffsim.diffsim(image_A=ref_path,
image_B=image1_path,
img_size=args.image_size,
prompt=prompt,
target_block=args.target_block,
target_layer=args.target_layer,
target_step=args.target_step,
ip_adapter=args.ip_adapter,
seed=args.seed,
device=device,
similarity=args.similarity)
diff_ac = diffsim.diffsim(image_A=ref_path,
image_B=image2_path,
img_size=args.image_size,
prompt=prompt,
target_block=args.target_block,
target_layer=args.target_layer,
target_step=args.target_step,
ip_adapter=args.ip_adapter,
seed=args.seed,
device=device,
similarity=args.similarity)
elif args.metric == 'diffsim_xl':
diff_ab = diffsim_xl_score.diffsim_score(ref_path, image1_path, args.image_size, prompt, args.target_block, args.target_layer, args.target_step, args.similarity, args.seed)
diff_ac = diffsim_xl_score.diffsim_score(ref_path, image2_path, args.image_size, prompt, args.target_block, args.target_layer, args.target_step, args.similarity, args.seed)
elif args.metric == 'dit':
diff_ab = diffsim_dit.diffsim_score(ref_path, image1_path, args.image_size, prompt, args.target_block, args.target_layer, args.target_step, args.similarity, args.seed)
diff_ac = diffsim_dit.diffsim_score(ref_path, image2_path, args.image_size, prompt, args.target_block, args.target_layer, args.target_step, args.similarity, args.seed)
elif args.metric == 'clip_i':
diff_ab = clip_score.clipi_score(load_image(ref_path), load_image(image1_path))[0]
diff_ac = clip_score.clipi_score(load_image(ref_path), load_image(image2_path))[0]
elif args.metric == 'clip_cross':
diff_ab = clip_score.clip_cross_score(load_image(ref_path), load_image(image1_path), args.target_layer)
diff_ac = clip_score.clip_cross_score(load_image(ref_path), load_image(image2_path), args.target_layer)
elif args.metric == 'clipfeats':
diff_ab = clip_score.clip_feature_score(load_image(ref_path), load_image(image1_path), args.target_layer)
diff_ac = clip_score.clip_feature_score(load_image(ref_path), load_image(image2_path), args.target_layer)
elif args.metric == 'dino' or args.metric == 'dinov1':
diff_ab = dino_score.dino_score(load_image(ref_path), load_image(image1_path))[0]
diff_ac = dino_score.dino_score(load_image(ref_path), load_image(image2_path))[0]
elif args.metric == 'dino_cross':
diff_ab = dino_score.dino_cross_score(load_image(ref_path), load_image(image1_path), args.target_layer)
diff_ac = dino_score.dino_cross_score(load_image(ref_path), load_image(image2_path), args.target_layer)
elif args.metric == 'dinofeats':
diff_ab = dino_score.dino_feature_score(load_image(ref_path), load_image(image1_path), args.target_layer)
diff_ac = dino_score.dino_feature_score(load_image(ref_path), load_image(image2_path), args.target_layer)
elif args.metric == 'cute':
diff_ab = cute_score("Crop-Feat", [load_image(ref_path)], [load_image(image1_path)])
diff_ac = cute_score("Crop-Feat", [load_image(ref_path)], [load_image(image2_path)])
elif args.metric == 'lpips':
diff_ab = lpips_score(process_image(load_image(ref_path)), process_image(load_image(image1_path))).item()
diff_ac = lpips_score(process_image(load_image(ref_path)), process_image(load_image(image2_path))).item()
elif args.metric == 'ensemble':
diff_ab = diffsim.diffsim(image_A=ref_path,
image_B=image1_path,
img_size=args.image_size,
prompt=prompt,
target_block=args.target_block,
target_layer=args.target_layer,
target_step=args.target_step,
ip_adapter=args.ip_adapter,
seed=args.seed,
device=device,
similarity=args.similarity)
diff_ac = diffsim.diffsim(image_A=ref_path,
image_B=image2_path,
img_size=args.image_size,
prompt=prompt,
target_block=args.target_block,
target_layer=args.target_layer,
target_step=args.target_step,
ip_adapter=args.ip_adapter,
seed=args.seed,
device=device,
similarity=args.similarity)
clip_ab = clip_score.clipi_score(load_image(ref_path), load_image(image1_path))[0]
clip_ac = clip_score.clipi_score(load_image(ref_path), load_image(image2_path))[0]
dino_ab = dino_score.dino_score(load_image(ref_path), load_image(image1_path))[0]
dino_ac = dino_score.dino_score(load_image(ref_path), load_image(image2_path))[0]
if args.metric == 'ensemble':
diff_corr = 0 if diff_ab < diff_ac else 1
clip_corr = 0 if clip_ab < clip_ac else 1
dino_corr = 0 if dino_ab < dino_ac else 1
if diff_corr + clip_corr + dino_corr >= 2:
correct_predictions += 1
else:
if args.metric in ["lpips", "dreamsim"] or ('diffsim' in args.metric and args.similarity == 'mse'):
if diff_ab < diff_ac:
correct_predictions += 1
elif diff_ab > diff_ac:
correct_predictions += 1
total_samples += 1
# if total_samples % 500 == 0:
# print(f"Total samples now: {total_samples}")
# print(f'Current accuracy: {correct_predictions / total_samples * 100}%')
print(f"Total samples now: {total_samples}")
print(f'Current accuracy: {correct_predictions / total_samples * 100}%')
if __name__ == "__main__":
args = arg_parse()
device = 'cuda'
ipref1(args, args.image_path, args.original_path, device)