-
Notifications
You must be signed in to change notification settings - Fork 0
/
Minibatch.py
68 lines (59 loc) · 2.51 KB
/
Minibatch.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
import numpy as np
import cv2
from GenerateLabel import From_dir_get_label
from PicMean import computeAllPicMean
class TrainData:
def __init__(self, minibatch = 8):
print('Init Train Data')
self._minibatch = minibatch
self._start = 0
#self._end = self._start + minibatch
def generateLabelsAPath(self, dir):
self._dataPaths, self._dataLabels, self._numClass = From_dir_get_label(dir)
print('generate Path And Label is finish !')
return self._dataPaths, self._dataLabels, self._numClass
def SetMean(self,mean):
self._mean = mean
def getNextMinibatch(self, start):
blobs = np.zeros((self._minibatch, 224, 224, 3), np.float32)
labels = np.zeros(self._minibatch, np.float32)
end = self._start + self._minibatch
if end >= len(self._dataLabels):
end = len(self._dataLabels)
# cv2.namedWindow('facePicture')
for index in range(self._start,end):
dataPath = self._dataPaths[index]
img = cv2.imread(dataPath)
img = cv2.resize(img, (224,224))
img -= self._mean
# cv2.imshow('facePicture',img)
# cv2.waitKey(10)
img = img.astype(np.float32, copy=False)
#减去均值
blobs[index,:,:,:] = img
labels[index] = self._dataLabels[index]
#cv2.destroyWindow('facePicture')
self._start += self._minibatch
return blobs, labels
def getRandomMinibatch(self):
blobs = np.zeros((self._minibatch, 224, 224, 3), np.float32)
labels = np.zeros(self._minibatch, np.float32)
randomIndex = np.random.randint(0,labels.shape[0], self._minibatch)
for i in range(randomIndex.shape[0]):
dataPath = self._dataPaths[randomIndex[i]]
img = cv2.imread(dataPath)
img = cv2.resize(img, (224, 224))
img -= self._mean
img = img.astype(np.float32, copy=False)
blobs[i, :, :, :] = img
labels[i] = self._dataLabels[randomIndex[i]]
return blobs, labels
#mean 88 67 61
if __name__ == '__main__':
trainData = TrainData()
dataPaths, dataLabels, numClass = trainData.generateLabelsAPath("G:\\DATESETS\\64_CASIA-FaceV5\\data")
trainData.SetMean(np.array([[88, 67, 61]], np.uint8))
for i in range(int(2500/16)):
blobs, labels = trainData.getNextMinibatch(i)
i += 16
print('Get All img and label !')