-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathFER.py
340 lines (267 loc) · 10.3 KB
/
FER.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
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
# -*- coding: utf-8 -*-
"""Facial.ipynb
Automatically generated by Colaboratory.
Original file is located at
https://colab.research.google.com/drive/1V8UyIoFp0Dfa4zItiu6pJakex8xNmgzQ
"""
# Commented out IPython magic to ensure Python compatibility.
from google.colab import drive
drive.mount('/content/gdrive')
# %cd /content/gdrive/My\ Drive/Projects/FER
import torchvision.transforms as transforms
import numpy as np
import h5py
import torch.utils.data as data
import torch
from torch.utils.data import DataLoader, Dataset
import torchvision
import torchvision.datasets as datasets
from PIL import Image
import matplotlib.pyplot as plt
import torch.nn as nn
import torch.nn.functional as F
import cv2
from google.colab.patches import cv2_imshow
class FER(data.Dataset):
def __init__(self, split='Training', transform=None):
self.transform = transform
self.split = split
self.data = h5py.File('./fer1.h5', 'r', driver='core')
if self.split == 'Training':
self.train_data = self.data['Training_pixel']
self.train_labels = self.data['Training_label']
self.train_data = np.asarray(self.train_data)
self.train_data = self.train_data.reshape((28709, 48, 48))
elif self.split == 'PublicTest':
self.PublicTest_data = self.data['PublicTest_pixel']
self.PublicTest_labels = self.data['PublicTest_label']
self.PublicTest_data = np.asarray(self.PublicTest_data)
self.PublicTest_data = self.PublicTest_data.reshape((3589, 48, 48))
else:
self.PrivateTest_data = self.data['PrivateTest_pixel']
self.PrivateTest_label = self.data['PrivateTest_label']
self.PrivateTest_data = np.asarray(self.PrivateTest_data)
self.PrivateTest_data = self.PrivateTest_data.reshape((3589, 48, 48))
def __len__(self):
if self.split == 'Training':
return len(self.train_data)
elif self.split == 'PublicTest':
return len(self.PublicTest_data)
else:
return len(self.PrivateTest_data)
def __getitem__(self, index):
if self.split == 'Training':
img, target = self.train_data[index], self.train_labels[index]
elif self.split == 'PublicTest':
img, target = self.PublicTest_data[index], self.PublicTest_labels[index]
else:
img, target = self.PrivateTest_data[index], self.PrivateTest_label[index]
img = img[:, :, np.newaxis]
img = np.concatenate((img, img, img), axis=2)
img = Image.fromarray(img)
if self.transform is not None:
img = self.transform(img)
return img, target
transform = transforms.Compose(
[transforms.RandomHorizontalFlip(),transforms.ToTensor(),
transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])
trainset = FER(split='Training', transform= transform)
transform1 = transforms.Compose(
[transforms.ToTensor(),
transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])
testset = FER(split='PublicTest', transform= transform1)
trainloader = torch.utils.data.DataLoader(trainset, batch_size=100, shuffle=True)
type(trainloader)
testloader = torch.utils.data.DataLoader(testset, batch_size=100, shuffle=False)
type(testloader)
for img, target in trainloader:
print(img.shape)
print(target)
break
img_data = img[0]
img_data.shape
np_image = img_data.numpy()
np_image = np.transpose(np_image, (1,2,0))
np_image = np_image*(0.5, 0.5, 0.5)+(0.5, 0.5, 0.5)
print(np_image.shape)
plt.figure(figsize = (2,2))
plt.imshow(np_image)
plt.show()
_, axs = plt.subplots(1, 8, figsize=(12, 12))
axs = axs.flatten()
for img, ax in zip(img, axs):
np_image = img.numpy()
np_image = np.transpose(np_image, (1,2,0))
np_image = np_image*(0.5, 0.5, 0.5)+(0.5, 0.5, 0.5)
ax.imshow(np_image)
plt.show()
class Net(nn.Module):
def __init__(self):
super(Net, self).__init__()
self.conv1 = nn.Conv2d(in_channels=3, out_channels=16, kernel_size=3)
self.pool = nn.MaxPool2d(kernel_size=(2,2), stride=2)
self.conv2 = nn.Conv2d(16,64,3)
self.conv3 = nn.Conv2d(64,128,3)
self.conv4 = nn.Conv2d(128,256,3)
self.fc1 = nn.Linear(256 * 1 * 1, 500)
self.fc2 = nn.Linear(500, 200)
self.fc3 = nn.Linear(200, 7)
def forward(self, x):
x = self.pool(F.relu(self.conv1(x)))
x = self.pool(F.relu(self.conv2(x)))
x = self.pool(F.relu(self.conv3(x)))
x = self.pool(F.relu(self.conv4(x)))
x = x.view(-1, 256 * 1 * 1)
x = F.relu(self.fc1(x))
x = F.relu(self.fc2(x))
x = self.fc3(x)
return(x)
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
print(device)
model = Net().to(device)
model
criteria = nn.CrossEntropyLoss()
optimizer = torch.optim.Adam(model.parameters(), lr=0.001)
loss_list = []
acc_list = []
val_list=[]
valoss_list = []
for e in range(100):
loss_net= 0
acc_net = 0
val_acc = 0
val_loss = 0
total = 0
total1 = 0
for images, labels in trainloader:
images = images.to(device)
labels = labels.to(device)
pred = model(images)
loss = criteria(pred,labels)
optimizer.zero_grad()
loss.backward()
optimizer.step()
_, pred = torch.max(pred.data,1)
loss_net += loss.item()
acc_net += pred.eq(labels.data).sum()
total += labels.size(0)
else:
with torch.no_grad(): #we dont want to update gradient from validation set
for images, labels in testloader:
images = images.to(device)
labels = labels.to(device)
pred = model(images)
loss_val = criteria(pred,labels)
_, pred = torch.max(pred.data,1)
val_loss += loss_val.item()
val_acc += pred.eq(labels.data).sum()
total1 += labels.size(0)
val_loss /= total1
valoss_list.append(val_loss)
val_acc = (int(val_acc.data)/total1)*100.0
val_list.append(val_acc)
loss_net /= total
acc_net = (int(acc_net.data)/total)*100.0
loss_list.append(loss_net)
acc_list.append(acc_net)
print("Epoch:",e,"loss: ", loss_net, "acc:", acc_net, "Val Acc: ", val_acc, "Val_loss: ", val_loss)
break
torch.save(model.state_dict(), "/content/gdrive/My Drive/Projects/FER/model")
plt.plot(loss_list, label ='Training Loss')
plt.plot(valoss_list, label = 'Validation loss')
plt.legend()
plt.plot(acc_list, label ='Training Acc')
plt.plot(val_list, label = 'Validation Acc')
plt.legend()
model.load_state_dict(torch.load("/content/gdrive/My Drive/Projects/FER/model"))
model.eval()
emotion_dict = {0: "Angry", 1: "Disgust", 2: "Fear", 3: "Happy", 4: "Sad", 5: "Surprise", 6: "Neutral"}
cap = cv2.VideoCapture('/content/gdrive/My Drive/Projects/FER/Final.mp4')
im_vd = []
num = 0
while True:
ret, frame = cap.read()
if not ret:
print("here")
continue
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_default.xml')
faces = face_cascade.detectMultiScale(gray, 1.3, 5)
for (x, y, w, h) in faces:
cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 1)
roi_gray = gray[y:y + h, x:x + w]
cropped_img = cv2.resize(roi_gray, (48, 48))
cropped_img = np.expand_dims(cropped_img, -1)
cropped_img = np.concatenate((cropped_img, cropped_img, cropped_img), axis=2)
cropped_img = Image.fromarray(cropped_img)
cropped_img = transform(cropped_img)
cropped_img = cropped_img[np.newaxis, :, :]
cropped_img = cropped_img.to(device)
pred = model(cropped_img)
pred = pred.cpu().data.numpy()
pred1 = np.argmax(pred[0])
pred_emo = emotion_dict[pred1]
cv2.putText(frame, emotion_dict[int(np.argmax(pred[0]))], (x, y), cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 255, 0), 1, cv2.LINE_AA)
im_vd.append(frame)
#cv2_imshow(resized_img)
num = num+1
if num == 150:#works on 100
break
size=(1080,720)
out = cv2.VideoWriter('/content/gdrive/My Drive/Projects/FER/project8.mp4',cv2.VideoWriter_fourcc(*'MJPG'), 15, size)
for i in range(len(im_vd)):
out.write(im_vd[i])
#cv2_imshow(im_vd[i])
print('end')
out.release()
cv2.destroyAllWindows
cap.release()
cv2.destroyAllWindows()
face_haar_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_default.xml')
cap=cv2.VideoCapture('gdrive/My Drive/Projects/FER/Final.mp4')
#cap = cv2.VideoCapture(0) #for capturing a video stream using camera. Does not work on Colab
im_vd =[]
num = 0
while True:
ret,test_img=cap.read()# captures frame and returns boolean value and captured image
if not ret:
print("here")
continue
gray_img= cv2.cvtColor(test_img, cv2.COLOR_BGR2GRAY)
faces_detected = face_haar_cascade.detectMultiScale(gray_img, 1.32, 5)
for (x,y,w,h) in faces_detected:
cv2.rectangle(test_img,(x,y),(x+w,y+h),(255,0,0),thickness=7)
roi_gray=gray_img[y:y+w,x:x+h]#cropping region of interest i.e. face area from image
roi_gray=cv2.resize(roi_gray,(48,48))
img_pixels = Image2.img_to_array(roi_gray)
img_pixels = np.expand_dims(img_pixels, axis = 0)
#img_pixels /= 255
img_pixels = np.concatenate((img_pixels,img_pixels,img_pixels),axis =3) #concatenate along channel dimension
img_pixels = img_pixels.astype(np.uint8)
img_pixels = img_pixels.reshape((48,48,3))
img_pixels = Image1.fromarray(img_pixels)
img_pixels = transform(img_pixels)
#img_pixels = img_pixels.transpose(2,0)
img_pixels = img_pixels[np.newaxis,:,:,:]
img_pixels = img_pixels.to(device)
predictions = model(img_pixels)
#to find max indexed array
pred = predictions.cpu().data.numpy()
max_index = np.argmax(pred[0])
emotions = ('angry', 'disgust', 'fear', 'happy', 'sad', 'surprise', 'neutral' )
predicted_emotion = emotions[max_index]
cv2.putText(test_img, predicted_emotion, (int(x), int(y)), cv2.FONT_HERSHEY_SIMPLEX, 1, (0,0,255), 2)
#resized_img = cv2.resize(test_img, (1000, 700))
print(test_img.shape)
im_vd.append(test_img)
#cv2_imshow(resized_img)
num = num+1
if num == 100:#works on 100
break
size=(1080,720)
out = cv2.VideoWriter('gdrive/My Drive/Projects/FER/project3.mp4',cv2.VideoWriter_fourcc(*'MJPG'), 15, size)
for i in range(len(im_vd)):
out.write(im_vd[i])
cv2_imshow(test_img)
out.release()
cv2.destroyAllWindows