-
Notifications
You must be signed in to change notification settings - Fork 0
/
rock_c3_infer2.py
158 lines (126 loc) · 5.76 KB
/
rock_c3_infer2.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
"""
eureka_infer.py
Zhiang Chen, Jan 6 2020
eureka data inference
"""
import transforms as T
from engine import train_one_epoch, evaluate
import utils
import torch
from rock_c3 import Dataset
from model import get_rock_model_instance_segmentation
import os
from shutil import copyfile
import pickle
import numpy as np
from model import visualize_result
from model import visualize_pred
class ToTensor(object):
def __call__(self, image, target):
# image = F.to_tensor(image).float()
image = torch.from_numpy(image / 255.0).float()
image = image.permute((2, 0, 1))
return image, target
def get_transform(train):
transforms = []
transforms.append(ToTensor()) # torchvision.transforms.functional is a garbage, sorry guys
return T.Compose(transforms)
def test_performance(model, data, device, path):
model.load_state_dict(torch.load(path + "/epoch_0039.param"))
evaluate(model, data, device=device)
model.load_state_dict(torch.load(path + "/epoch_0033.param"))
evaluate(model, data, device=device)
model.load_state_dict(torch.load(path + "/epoch_0030.param"))
evaluate(model, data, device=device)
model.load_state_dict(torch.load(path + "/epoch_0025.param"))
evaluate(model, data, device=device)
def get_mean_std(input_channel, image_mean, image_std):
if input_channel == 8:
return image_mean, image_std
elif input_channel == 3:
return image_mean[:3], image_std[:3]
elif input_channel == 5:
return image_mean[:5], image_std[:5]
elif input_channel == 6:
return image_mean[:3] + image_mean[-3:], image_std[:3] + image_mean[-3:]
elif input_channel == 4:
return image_mean[:3] + [np.mean(image_mean[-3:]).tolist()], image_std[:3] + [np.mean(image_std[-3:]).tolist()]
elif input_channel == 'dem':
return image_mean[-3:], image_std[-3:]
if __name__ == '__main__':
# train on the GPU or on the CPU, if a GPU is not available
device = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu')
device = torch.device('cuda:0')
# our dataset has three classes only - background, non-damaged, and damaged
num_classes = 2
input_c = 3
dataset_infer = Dataset("./datasets/C3/rgbd/", transforms=get_transform(train=False), include_name=True, input_channel=input_c)
dataset_test = Dataset("./datasets/C3_test/all_rocks/", transforms=get_transform(train=False), include_name=False, input_channel=input_c)
dataset = Dataset("./datasets/C3/aug/", transforms=get_transform(train=True), input_channel=input_c)
image_mean, image_std, _, _ = dataset.imageStat()
# image_mean = [0.23924888725523394, 0.2180423480395164, 0.2118836715688813, 0.26721142156890876, 0.32996910784324385, 0.1461123186277879, 0.5308107499991753, 0.28652559313771186]
# image_std = [0.1459739643338365, 0.1311105424825076, 0.12715888419418298, 0.149469170605332, 0.15553466224696225, 0.10533129832132752, 0.24088403135495345, 0.24318892151508417]
image_mean, image_std = get_mean_std(input_c, image_mean, image_std)
data_loader_test = torch.utils.data.DataLoader(
dataset_test, batch_size=1, shuffle=False, num_workers=2,
collate_fn=utils.collate_fn)
mask_rcnn = get_rock_model_instance_segmentation(num_classes, input_channel=input_c, image_mean=image_mean, image_std=image_std)
# move model to the right device
mask_rcnn.to(device)
mask_rcnn.eval()
# test_performance(mask_rcnn, data_loader_test, device, "trained_param_8")
instances = []
'''
init_epoch = 10
num_epochs = 25
for epoch in range(init_epoch, init_epoch + num_epochs):
"""
# train for one epoch, printing every 10 iterations
train_one_epoch(mask_rcnn, optimizer, data_loader, device, epoch, print_freq=100)
# update the learning rate
lr_scheduler.step()
"""
print("trained_param_c3_4/epoch_00%02d.param" % epoch)
mask_rcnn.load_state_dict(torch.load("trained_param_c3_4/epoch_00%02d.param" % epoch))
# evaluate on the test dataset
evaluate(mask_rcnn, data_loader_test, device=device)
'''
mask_rcnn.load_state_dict(torch.load("trained_param_c3_3/epoch_0018.param"))
#mask_rcnn.load_state_dict(torch.load("trained_param_c3_3/epoch_0002.param"))
f = 0
for i, data in enumerate(dataset_infer):
print(i)
image, target = data
pred = mask_rcnn(image.unsqueeze(0).to(device))[0]
boxes = pred['boxes'].to("cpu").detach().numpy()
labels = pred['labels'].to("cpu").detach().numpy()
scores = pred['scores'].to("cpu").detach().numpy()
masks = pred['masks'].to("cpu").detach().numpy()
image_name = target['image_name']
result = {}
result['bb'] = boxes
result['labels'] = labels
result['scores'] = scores
result['mask'] = masks
result['image_name'] = image_name
result['coord'] = [int(i)*390 for i in image_name.split('/')[-1].split('.')[0].split('_')]
nm = masks.shape[0]
for i in range(nm):
rock = {}
rock['bb'] = boxes[i]
rock['mask'] = masks[i, 0, :, :]
rock['score'] = scores[i]
rock['coord'] = [int(i)*390 for i in image_name.split('/')[-1].split('.')[0].split('_')]
instances.append(rock)
#visualize_result(mask_rcnn, data)
#visualize_pred(image, pred)
if len(instances) >= 30000:
name = "./datasets/C3/rocks_c3_3_18_%02d.pickle" % f
f += 1
with open(name, 'wb') as filehandle:
pickle.dump(instances, filehandle)
instances = []
name = "./datasets/C3/rocks_c3_3_18_%02d.pickle" % f
with open(name, 'wb') as filehandle:
pickle.dump(instances, filehandle)
#'''