-
Notifications
You must be signed in to change notification settings - Fork 7
/
main_webvision.py
311 lines (263 loc) · 15.3 KB
/
main_webvision.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
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
'''
Update 2022.11.5
To ensure a direct minimal reproduction applicable, I removed the codes on imagenet evaluation. [Marked with *]
To evaluate on iamgenet, please modify the code yourself
'''
import argparse
import torchvision.transforms as transforms
import wandb
from torch.optim import SGD
from torch.utils.data import Subset
from tqdm import tqdm
from datasets.dataloader_webvision import miniwebvision_dataset, imagenet_dataset
from models.inceptionresnetv2 import InceptionResNetV2
from utils import *
parser = argparse.ArgumentParser('Train with Webvision dataset')
parser.add_argument('--dataset_path', default='~/WebVision', help=f'dataset path')
#*************************************************************************************************************#
# parser.add_argument('--imagenet_path', default='~/ImageNet', help=f'dataset_path for imagenet evaluation')
#*************************************************************************************************************#
# model settings
parser.add_argument('--lambda_fc', default=1, type=float, metavar='N', help='weight of unlabeled data (default: 1)')
parser.add_argument('--theta_s', default=1.0, type=float, help='Initial threshold for voted correct samples (default: 1.0)')
parser.add_argument('--theta_r', default=0.95, type=float, help='threshold for relabel samples (default: 0.8)')
parser.add_argument('--k', default=200, type=int, help='neighbors for soft-voting (default: 200)')
# train settings
parser.add_argument('--epochs', default=150, type=int, metavar='N', help='number of total epochs to run (default: 300)')
parser.add_argument('--batch_size', default=32, type=int, help='mini-batch size (default: 32)')
parser.add_argument('--lr', default=0.01, type=float, help='initial learning rate (default: 0.1)')
parser.add_argument('--momentum', default=0.9, type=float, metavar='M', help='momentum of SGD solver (default: 0.9)')
parser.add_argument('--weight_decay', default=1e-4, type=float, help='weight decay (default: 5e-4)')
parser.add_argument('--seed', default=3047, type=int, help='seed for initializing training. (default: 3047)')
parser.add_argument('--parallel', default=0, action='store_true', help='Multi-GPU training (default: False)')
parser.add_argument('--gpuid', default='0', type=str, help='Selected GPU (default: "0")')
parser.add_argument('--entity', type=str, help='Wandb user entity')
parser.add_argument('--run_path', type=str, help='run path containing all results')
def train(labeled_trainloader, modified_label, all_trainloader, encoder, classifier, proj_head, pred_head, optimizer, epoch, args):
encoder.train()
classifier.train()
proj_head.train()
pred_head.train()
xlosses = AverageMeter('xloss')
ulosses = AverageMeter('uloss')
labeled_train_iter = iter(labeled_trainloader)
all_bar = tqdm(all_trainloader)
for batch_idx, ([inputs_u1, inputs_u2], _, _) in enumerate(all_bar):
try:
# [inputs_x1, inputs_x2], labels_x, index = labeled_train_iter.next()
[inputs_x1, inputs_x2], labels_x, index = next(labeled_train_iter)
except:
labeled_train_iter = iter(labeled_trainloader)
# [inputs_x1, inputs_x2], labels_x, index = labeled_train_iter.next()
[inputs_x1, inputs_x2], labels_x, index = next(labeled_train_iter)
# cross-entropy training with mixup
batch_size = inputs_x1.size(0)
inputs_x1, inputs_x2 = inputs_x1.cuda(), inputs_x2.cuda()
labels_x = modified_label[index]
targets_x = torch.zeros(batch_size, args.num_classes, device=inputs_x1.device).scatter_(1, labels_x.view(-1, 1), 1)
l = np.random.beta(0.5, 0.5)
l = max(l, 1 - l)
all_inputs_x = torch.cat([inputs_x1, inputs_x2], dim=0)
all_targets_x = torch.cat([targets_x, targets_x], dim=0)
idx = torch.randperm(all_inputs_x.size()[0])
input_a, input_b = all_inputs_x, all_inputs_x[idx]
target_a, target_b = all_targets_x, all_targets_x[idx]
mixed_input = l * input_a + (1 - l) * input_b
mixed_target = l * target_a + (1 - l) * target_b
logits = classifier(encoder(mixed_input))
Lce = -torch.mean(torch.sum(F.log_softmax(logits, dim=1) * mixed_target, dim=1))
# optional feature-consistency
inputs_u1, inputs_u2 = inputs_u1.cuda(), inputs_u2.cuda()
feats_u1 = encoder(inputs_u1)
feats_u2 = encoder(inputs_u2)
f, h = proj_head, pred_head
z1, z2 = f(feats_u1), f(feats_u2)
p1, p2 = h(z1), h(z2)
Lfc = D(p2, z1)
loss = Lce + args.lambda_fc * Lfc
xlosses.update(Lce.item())
ulosses.update(Lfc.item())
all_bar.set_description(
f'Train epoch {epoch} LR:{optimizer.param_groups[0]["lr"]} Labeled loss: {xlosses.avg:.4f} Unlabeled loss: {ulosses.avg:.4f}')
optimizer.zero_grad()
loss.backward()
optimizer.step()
logger.log({'ce loss': xlosses.avg, 'fc loss': ulosses.avg})
def test(testloader, encoder, classifier, epoch):
encoder.eval()
classifier.eval()
# accuracy = AverageMeter('accuracy')
accs1 = AverageMeter('accs1')
accs5 = AverageMeter('accs5')
data_bar = tqdm(testloader)
with torch.no_grad():
for i, (data, label, _) in enumerate(data_bar):
data, label = data.cuda(), label.cuda()
feat = encoder(data)
res = classifier(feat)
acc1, acc5 = accuracy(res, label, [1, 5])
accs1.update(acc1.item(), data.size(0))
accs5.update(acc5.item(), data.size(0))
data_bar.set_description(f'Test epoch {epoch}: Accuracy1@{accs1.avg:.4f} Accuracy5@{accs5.avg:.4f}')
return accs1.avg, accs5.avg
def evaluate(dataloader, encoder, classifier, args, noisy_label):
encoder.eval()
classifier.eval()
feature_bank = []
prediction = []
with torch.no_grad():
# generate feature bank
for (data, target, index) in tqdm(dataloader, desc='Feature extracting'):
data = data.cuda()
feature = encoder(data)
feature_bank.append(feature)
res = classifier(feature)
prediction.append(res)
feature_bank = F.normalize(torch.cat(feature_bank, dim=0), dim=1)
################################### sample relabelling ###################################
prediction_cls = torch.softmax(torch.cat(prediction, dim=0), dim=1)
his_score, his_label = prediction_cls.max(1)
print(f'Prediction track: mean: {his_score.mean()} max: {his_score.max()} min: {his_score.min()}')
conf_id = torch.where(his_score > args.theta_r)[0]
modified_label = torch.clone(noisy_label).detach()
modified_label[conf_id] = his_label[conf_id]
################################### sample selection ###################################
prediction_knn = weighted_knn(feature_bank, feature_bank, modified_label, args.num_classes, args.k, 10) # temperature in weighted KNN
vote_y = torch.gather(prediction_knn, 1, modified_label.view(-1, 1)).squeeze()
vote_max = prediction_knn.max(dim=1)[0]
right_score = vote_y / vote_max
clean_id = torch.where(right_score >= args.theta_s)[0]
noisy_id = torch.where(right_score < args.theta_s)[0]
return clean_id, noisy_id, modified_label
def main():
args = parser.parse_args()
seed_everything(args.seed)
if args.run_path is None:
args.run_path = f'Dataset(webvision_Model({args.theta_r}_{args.theta_s})'
global logger
logger = wandb.init(project='webvision', entity=args.entity, name=args.run_path)
logger.config.update(args)
os.environ['CUDA_VISIBLE_DEVICES'] = args.gpuid
# generate noisy dataset with our transformation
if not os.path.isdir(f'webvision'):
os.mkdir(f'webvision')
if not os.path.isdir(f'webvision/{args.run_path}'):
os.mkdir(f'webvision/{args.run_path}')
args.num_classes = 50
################################ Model initialization ###########################################
encoder = InceptionResNetV2(args.num_classes)
dim = encoder.last_linear.in_features
encoder.last_linear = torch.nn.Identity()
classifier = torch.nn.Linear(dim, args.num_classes)
proj_head = torch.nn.Sequential(torch.nn.Linear(dim, 256),
torch.nn.BatchNorm1d(256),
torch.nn.ReLU(),
torch.nn.Linear(256, 256))
pred_head = torch.nn.Sequential(torch.nn.Linear(256, 256),
torch.nn.BatchNorm1d(256),
torch.nn.ReLU(),
torch.nn.Linear(256, 256))
encoder.cuda()
classifier.cuda()
proj_head.cuda()
pred_head.cuda()
if args.parallel:
encoder = torch.nn.DataParallel(encoder).cuda()
classifier = torch.nn.DataParallel(classifier).cuda()
proj_head = torch.nn.DataParallel(proj_head).cuda()
pred_head = torch.nn.DataParallel(pred_head).cuda()
############################# Dataset initialization ##############################################
normalize = transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225))
weak_transform = transforms.Compose([
transforms.Resize(320),
transforms.RandomResizedCrop(299),
transforms.RandomHorizontalFlip(),
transforms.ToTensor(),
normalize])
none_transform = transforms.Compose([
transforms.Resize(320),
transforms.CenterCrop(299),
transforms.ToTensor(),
normalize]) # no augmentation
strong_transform = transforms.Compose([transforms.Resize(320),
transforms.RandomResizedCrop(299),
transforms.RandomHorizontalFlip(),
ImageNetPolicy(),
transforms.ToTensor(),
normalize])
test_transform = none_transform
eval_data = miniwebvision_dataset(root_dir=args.dataset_path, transform=weak_transform, dataset_mode='train',num_class=args.num_classes)
test_data = miniwebvision_dataset(root_dir=args.dataset_path, transform=test_transform, dataset_mode='test', num_class=args.num_classes)
train_data = miniwebvision_dataset(root_dir=args.dataset_path, transform=KCropsTransform(strong_transform, 2),
dataset_mode='train', num_class=args.num_classes)
all_data = miniwebvision_dataset(root_dir=args.dataset_path, transform=MixTransform(strong_transform, weak_transform, 1),
dataset_mode='train', num_class=args.num_classes)
noisy_label = torch.tensor(eval_data.train_labels).cuda()
eval_loader = torch.utils.data.DataLoader(eval_data, batch_size=args.batch_size * 10, shuffle=False, pin_memory=True, num_workers=4)
test_loader = torch.utils.data.DataLoader(test_data, batch_size=args.batch_size, shuffle=False, pin_memory=True, num_workers=4)
all_loader = torch.utils.data.DataLoader(all_data, batch_size=args.batch_size, num_workers=4, shuffle=True)
# *************************************************************************************************************#
# to evaluate on imagenet_datset
# test_data2 = imagenet_dataset(transform=test_transform)
# test_loader2 = torch.utils.data.DataLoader(test_data2, batch_size=args.batch_size, shuffle=False, pin_memory=True, num_workers=4)
# *************************************************************************************************************#
#################################### Training initialization #######################################
optimizer = SGD([{'params': encoder.parameters()}, {'params': classifier.parameters()}, {'params': proj_head.parameters()},
{'params': pred_head.parameters()}], lr=args.lr, weight_decay=args.weight_decay, momentum=args.momentum)
scheduler = torch.optim.lr_scheduler.MultiStepLR(optimizer, milestones=[50, 100], gamma=0.1)
acc_logs = open(f'webvision/{args.run_path}/acc.txt', 'w')
stat_logs = open(f'webvision/{args.run_path}/stat.txt', 'w')
save_config(args, f'{args.run_path}')
# best_acc1img = 0
# best_acc5img = 0
best_acc1web = 0
best_acc5web = 0
print('Train args: \n', args)
################################ Training loop ###########################################
for i in range(args.epochs):
clean_id, noisy_id, modified_label = evaluate(eval_loader, encoder, classifier, args, noisy_label)
print(f'Epoch [{i}/{args.epochs}]: clean samples_1: {len(clean_id)}, noisy samples_1: {len(noisy_id)}')
labeled_data = Subset(train_data, clean_id.cpu())
sampler = ClassBalancedSampler(labels=modified_label[clean_id], num_classes=args.num_classes)
labeled_loader = torch.utils.data.DataLoader(labeled_data, batch_size=args.batch_size, sampler=sampler, num_workers=4)
xloss, uloss = train(labeled_loader, modified_label, all_loader, encoder, classifier, proj_head, pred_head, optimizer, i, args)
stat_logs.write(f'Epoch [{i}/{args.epochs}]: clean samples_1: {len(clean_id)}, noisy samples_1: {len(noisy_id)} \n')
stat_logs.flush()
cur_acc1web, cur_acc5web = test(test_loader, encoder, classifier, i)
scheduler.step()
# *************************************************************************************************************#
# cur_acc1img, cur_acc5img = test(test_loader2, encoder, classifier, i)
# logger.log({'xloss': xloss, 'uloss': uloss, 'imgnet_acc1': cur_acc1img, 'imgnet_acc5': cur_acc5img, 'web_acc1': cur_acc1web, 'web_acc5': cur_acc5web})
# if cur_acc5img > best_acc5img:
# best_acc5img = cur_acc5img
# if cur_acc1img > best_acc1img:
# best_acc1img = cur_acc1img
# *************************************************************************************************************#
logger.log({'xloss': xloss, 'uloss': uloss, 'web_acc1': cur_acc1web, 'web_acc5': cur_acc5web})
if cur_acc5web > best_acc5web:
best_acc5web = cur_acc5web
if cur_acc1web > best_acc1web:
best_acc1web = cur_acc1web
save_checkpoint({
'cur_epoch': i,
'classifier': classifier.state_dict(),
'encoder': encoder.state_dict(),
'proj_head': proj_head.state_dict(),
'pred_head': pred_head.state_dict(),
'optimizer': optimizer.state_dict(),
}, filename=f'webvision/{args.run_path}/best_acc.pth.tar')
acc_logs.write(
f'Epoch [{i}/{args.epochs}]: web Best accuracy@1#:{best_acc1web}! Current accuracy@1#:{cur_acc1web}! Best accuracy@5#:{best_acc5web}! Current accuracy@5#:{cur_acc5web}\n')
# acc_logs.write(
# f'Epoch [{i}/{args.epochs}]: img Best accuracy@1#:{best_acc1img}! Current accuracy@1#:{cur_acc1img}! Best accuracy@5#:{best_acc5img}! Current accuracy@5#:{cur_acc5img}\n')
acc_logs.flush()
save_checkpoint({
'cur_epoch': i,
'classifier': classifier.state_dict(),
'encoder': encoder.state_dict(),
'proj_head': proj_head.state_dict(),
'pred_head': pred_head.state_dict(),
'optimizer': optimizer.state_dict(),
}, filename=f'webvision/{args.run_path}/last.pth.tar')
if __name__ == '__main__':
main()