-
Notifications
You must be signed in to change notification settings - Fork 3
/
trainer.py
199 lines (185 loc) · 6.57 KB
/
trainer.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
import numpy as np
import pdb
import cv2
print 'in import 1'
print 'in import 2'
from sklearn import svm
# you didnt have ti import openCV?
print 'in import'
import inspect
def lineno():
"""Returns the current line number in our program."""
return inspect.currentframe().f_back.f_lineno
# if name == '__main__':
# print "hello, this is line number", lineno()
print lineno()
videoLocation = '/Users/me/Desktop/MITDentalData/Videos/'
side = ['l','r']
# see about output array
print lineno()
output = [0,0,0,0,1,1,1,1]
svm_input = []
print lineno()
for j in range (0,2):
print lineno()
for i in range (1,4):
print "i = ", i, "j = ", j, " 1"
videoName = videoLocation+str(i)+side[j]+".mov"
videoNameDiff = videoLocation+str(i)+side[j]+"diff.mov"
cam = cv2.VideoCapture(videoName)
frame_limit = int(cam.get(cv2.cv.CV_CAP_PROP_FRAME_COUNT))
k, l=0, 0
poi = [4,8]
pdb.set_trace()
feature_vector = np.zeros((2,2,frame_limit),dtype=np.int)
# The error is in the line above
print "videoNAme is ", videoName
# Show thme the names of the npy files
read_npy = np.load(videoName+'.npy')
read_npy[:,0] = read_npy[:,0] - read_npy[52,0]
read_npy[:,1] = read_npy[:,1] - read_npy[52,1]
while k < frame_limit:
for l in range (0,len(poi)-1):
feature_vector[l][0][k] = read_npy[poi[l+1],0,j] - read_npy[poi[l],0,j]
feature_vector[l][1][k] = read_npy[poi[l+1],1,j] - read_npy[poi[l],1,j]
# The problem is that you are trying to access poi[2] when l=1, which does not exist since poi is a 2 member list, hence there is only poi[0] and poi[1]
np.save(videoNameDiff, feature_vector)
svm_input.append(feature_vector.tolist())
print "i = ", i, "j = ", j
# Oh god!!!! this is horrible. please open your terminal and and run it, i cant do it
i = i + 1
j = j + 1
clf = svm.SVC();
clf.fit(svm_input[1:6], output[1:6])
output = clf.predict_all(svm_input[:])
print output
# for i in range (1,15):
# for j in range (0,3):
# videoName = videoLocation+str(i)+side[j]+".mov"
# if (os.path.isfile(videoName)==False):
# print 'Did not find ',videoName
# videoName = videoLocation+str(i)+side[j+1]+".mov"
# if (os.path.isfile(videoName)==False):
# print 'Did not find ',videoName
# videoName = videoLocation+str(i)+side[j+2]+".mov"
# if (os.path.isfile(videoName)==False):
# pass
#
# cam = cv2.VideoCapture(videoName)
# frame_limit = int(cam.get(cv2.cv.CV_CAP_PROP_FRAME_COUNT))
# k=0
# while k<frame_limit:
# ret, img = cam.read()
# cv2.imwrite('temporary_image.png',img)
# directoryLocation = os.path.dirname(os.path.abspath(__file__))
# imageLocation = directoryLocation + '/temporary_image.png'
# img,points = landmark_locator(imageLocation, width=500, height=600, fps=10)
# np.save(videoLocation+str(i)+side[j]+".mov", points)
# print 'Generated .npy for ',videoLocation+str(i)+side[j]+".mov", 'frame',k
# k = k+1
#
#
# j = j+1
# i=i+1
#
# if __name__ == '__main__':
# frame_limit = 75
# feature_vector = np.zeros((4,2,frame_limit-1),dtype=np.int)
# directoryPath = "/Users/me/Desktop/MITDentalData/Videos/"
# array_name = [
# "1l.mov",
# "2l.mov",
# "3l.mov",
# "4l.mov",
# "1r.mov",
# "2r.mov",
# "3r.mov",
# "4r.mov"
# ]
#
# np.save(array_name[0], feature_vector)
#
# # array_name = [
# # "L1.mov",
# # "L2.mov",
# # "L3.mov",
# # "L4.mov",
# # "L8.mov",
# # "L9.mov",
# # "L10.mov",
# # "R1.mov",
# # "R2.mov",
# # "R3.mov",
# # "R4.mov",
# # "R5.mov",
# # "R6.mov",
# # "R7.mov"
# # ]
# # array_name_diff = [
# # "L1_diff.mov",
# # "L2_diff.mov",
# # "L3_diff.mov",
# # "L4_diff.mov",
# # "L8_diff.mov",
# # "L9_diff.mov",
# # "L10_diff.mov",
# # "R1_diff.mov",
# # "R2_diff.mov",
# # "R3_diff.mov",
# # "R4_diff.mov",
# # "R5_diff.mov",
# # "R6_diff.mov",
# # "R7_diff.mov"
# #
# # ]
# output = [0,0,0,0,1,1,1,1]
# svm_input = []
# poi = [5, 7, 59, 65]
#
# # store_x_values = []
# # store_y_values = []
# for loop_iterator in range(8):
# print loop_iterator
# read_npy = np.load(directoryPath+array_name[loop_iterator] + '.npy')
# print loop_iterator, array_name[loop_iterator]
# j = 0
#
# while j < frame_limit-1:
# for i in range (0,4):
# # if j==179:
# # feature_vector[i][0][j] = read_npy[i,0,j] #- read_npy[i,0,j+1]
# # feature_vector[i][1][j] = read_npy[i,1,j] #- read_npy[i,1,j+1]
# # else:
# # store_x_values = read_npy
# feature_vector[i][0][j] = read_npy[poi[i],0,j]
# feature_vector[i][1][j] = read_npy[poi[i],1,j]
# # np.save(array_name_diff[loop_iterator], feature_vector)
# # print j
# j = j+1
#
# # fourier_values = np.zeros((77,2,frame_limit-1))
# # var = np.zeros(77)
# # mean = np.zeros(77)
# # third_moment = np.zeros(77)
# # svm_input_temp = np.zeros(4*2)
# # for i in range(0,4):
# # # var[i] = np.var(feature_vector[i,0,:])
# # # mean[i] = np.mean(feature_vector[i,0,:])
# # # third_moment[i] = sum((feature_vector[i,0,:] - np.mean(feature_vector[i,0,:]))**3)/len(feature_vector[i,0,:])
# # # svm_input_temp[2*i] = var[i]
# # # svm_input_temp[2*i + 1] = third_moment[i]
# # svm_input_temp[i,0] = feature_vector[i][0]
# # svm_input_temp[i,1] = feature_vector[i][1]
# # svm_input.append(svm_input_temp.tolist())
# svm_input.append(feature_vector.tolist())
# # fourier_values[i,0,:] = np.fft.fft(np.sqrt(feature_vector[i,0,:]**2 + feature_vector[i,1,:]**2))
# # fourier_values[i,1,:] = np.fft.fft(np.arctan2(feature_vector[i,1,:], feature_vector[i,0,:]))
# # svm_input = zip(var, third_moment)
# # pdb.set_trace()
# # X = [[0, 0], [1, 1]]
# # Y = [0, 1]
# clf = svm.SVC()
# clf.fit(svm_input[5:11], output[5:11])
# clf.predict(svm_input[:])
#
# pdb.set_trace()