-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathutil.py
161 lines (147 loc) · 6.57 KB
/
util.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
import glob
import json
import dataset
import torch.nn as nn
import torch
from torchvision import transforms
import numpy as np
from torch.nn import functional as F
from PIL import Image
import torch.optim as optim
from myNetwork import network
from torch.utils.data import DataLoader
from torchvision.transforms import functional
from torch.autograd import Variable
from Dataset import Dataset
import os
import cv2
import time
import h5py
import matplotlib.pyplot as plt
plt.switch_backend('agg')
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
def get_one_hot(target,num_class):
one_hot=torch.zeros(target.shape[0],num_class).to(device)
one_hot=one_hot.scatter(dim=1,index=target.long().view(-1,1),value=1.)
return one_hot
class EoCNet:
def __init__(self,numclass,feature_extractor,batch_size,task_size,memory_size,epochs,learning_rate):
super(EoCNet, self).__init__()
self.epochs=epochs
self.learning_rate=learning_rate
self.model = network(feature_extractor)
self.exemplar_set = []
self.exemplar_set_gt=[]
self.class_mean_set = []
self.numclass = numclass
self.increntmal_phase= 0
self.train_list=list()
self.val_list = list()
self.batchsize = batch_size
self.memory_size=memory_size
self.task_size=task_size
self.workers=4
self.train_loader=None
self.test_loader=None
self.train_dataset=[]
self.image_list = list()
self.label_list = list()
self.class_name = ['others', 'IMG', 'jujube', 'cherry', 'tulip', 'chicken', 'vehicle']
self.transform = transforms.Compose([
transforms.Resize([400,400]),
transforms.ToTensor(),
transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225))])
def _test(self, test_loader,flag,model):
model.eval()
mae_final = 0
mse_final = 0
# accuracy_final =0
mae = 0
mae_total = 0
mse_total = 0
class_num=0
# correct=0
if flag == 'test':
num = {0: [0, 182], 1: [0, 248, 430], 2: [0, 182, 364, 612], 3: [0, 215, 397, 645, 827],
4: [0, 182, 364, 612, 794, 1009], 5: [0, 176, 358, 606, 788, 1003, 1185]
}
elif flag == 'val':
num = {0: [0, 59], 1: [0, 50, 109], 2: [0, 50, 109, 159], 3: [0, 40, 99, 149, 199],
4: [0, 40, 99, 149, 199, 239], 5: [0, 44, 103, 153, 203, 243, 283]
}
bottom=0
upper=1
all_test_dataset_img=test_loader.dataset.image_list
all_test_dataset_label=test_loader.dataset.label_list
all_test_dataset_target=test_loader.dataset.target
for index_ in range(len(num[self.increntmal_phase])-1):
#correct = 0
mae = 0
mse = 0
test_loader.dataset.image_list = all_test_dataset_img[num[self.increntmal_phase][bottom]:num[self.increntmal_phase][upper]]
test_loader.dataset.label_list = all_test_dataset_label[num[self.increntmal_phase][bottom]:num[self.increntmal_phase][upper]]
#test_loader.dataset.target = all_test_dataset_target[num[self.increntmal_phase][bottom]:num[self.increntmal_phase][upper]]
for i, (img, density) in enumerate(test_loader):
model=model.to(device)
density=density.to(device)
# target = target.to(device)
img=img.to(device)
with torch.no_grad():
output, cls, _, _ = model(img, 1)
for index in range(cls.shape[0]):
channel_num = torch.argmax(cls[index])
output = output[:, channel_num:channel_num + 1, :, :]
mae += abs(output.data.sum() - density.sum().type(torch.FloatTensor).cuda())
mse += ((output.data.sum()-density.sum()) ** 2).item()
mae_total += abs(output.data.sum() - density.sum().type(torch.FloatTensor).cuda())
mse_total += ((output.data.sum() - density.sum()) ** 2).item()
#accuracy = correct/len(test_loader.dataset.image_list)
mae = mae/len(test_loader.dataset.image_list)
mse = mse/len(test_loader.dataset.image_list)
mse = mse ** 0.5
print('class:%d,mae:%.2f，mse:%.2f' % (class_num,mae,mse))
#accuracy_final += accuracy
mae_final += mae
mse_final += mse
bottom = upper
upper = upper + 1
class_num+=1
#accuracy_final= accuracy_final/(self.increntmal_phase + 1)
mae_final = mae_final / (self.increntmal_phase + 1)
mse_final = mse_final / (self.increntmal_phase + 1)
test_loader.dataset.image_list = all_test_dataset_img
test_loader.dataset.label_list = all_test_dataset_label
test_loader.dataset.target = all_test_dataset_target
print(' * Average MAE :%.2f ' % (mae_final))
print(' * Average MSE :%.2f ' % (mse_final))
mae_total = mae_total / len(all_test_dataset_img)
mse_total = mse_total / len(all_test_dataset_img)
mse_total = mse_total ** 0.5
print(' ** MAE :%.2f ' % (mae_total))
print(' ** MSE :%.2f ' % (mse_total))
return mae_final
def test(self):
print('*begin test*')
if self.numclass>self.task_size:
self.model.Incremental_learning_weight(self.numclass)
checkpoint_best = torch.load(os.path.join('./checkpoint', 'checkpoint_best_'+str(self.increntmal_phase)+'.pth'))
self.model.load_state_dict(checkpoint_best['model'])
self.numclass += self.task_size
test_dataset = Dataset('test', self.increntmal_phase, None, None)
test_loader = DataLoader(test_dataset, batch_size=1, shuffle=False)
print('{0}th phase:the length of the test_dataset:{1}'.format(self.increntmal_phase, len(test_dataset)))
mae=self._test(test_loader,'test',self.model)
self.increntmal_phase += 1
def transform_image(self, image):
image = Image.open(image).convert('RGB')
height, width = image.size[1], image.size[0]
height = round(height / 16) * 16
width = round(width / 16) * 16
if height > 2000 or width > 2000:
height,width = 2000, 2000
image = image.resize((width, height), Image.BILINEAR)
else:
image = image.resize((width, height), Image.BILINEAR)
image = functional.to_tensor(image)
image = functional.normalize(image, [0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
return image