-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathgrad_cam.py
311 lines (251 loc) · 12.8 KB
/
grad_cam.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
import torch
import torch.nn as nn
import torch.nn.functional as F
import numpy as np
from dataset import IRDatasetFromNames, dataset_from_txt
def target_category_loss(x, category_index, nb_classes):
return torch.mul(x, F.one_hot(category_index, nb_classes))
def target_category_loss_output_shape(input_shape):
return input_shape
def normalize(x):
# utility function to normalize a tensor by its L2 norm
return x / (torch.sqrt(torch.mean(torch.square(x))) + 1e-5)
class ActivationsAndGradients:
""" Class for extracting activations and
registering gradients from targetted intermediate layers """
def __init__(self, model, target_layer):
self.model = model
self.gradients = []
self.activations = []
target_layer.register_forward_hook(self.save_activation)
target_layer.register_full_backward_hook(self.save_gradient)
def save_activation(self, module, input, output):
self.activations.append(output)
def save_gradient(self, module, grad_input, grad_output):
# Gradients are computed in reverse order
self.gradients = [grad_output[0]] + self.gradients
def __call__(self, x):
self.gradients = []
self.activations = []
return self.model(x)
class BaseCAM:
def __init__(self, model, target_layer, use_cuda=False):
self.model = model.eval()
self.target_layer = target_layer
self.cuda = use_cuda
if self.cuda:
self.model = model.cuda()
self.activations_and_grads = ActivationsAndGradients(self.model, target_layer)
def forward(self, input_img):
return self.model(input_img)
def get_cam_weights(self,
input_tensor,
target_category,
activations,
grads):
raise Exception("Not Implemented")
def get_loss(self, output, target_category):
# print(output.size())
return output[target_category]
def __call__(self, input_tensor, target_category=None):
if self.cuda:
input_tensor = input_tensor.cuda()
output = self.activations_and_grads(input_tensor)
if target_category is None:
output = output.squeeze()
target_category = np.argmax(output.cpu().data.numpy())
# print(output)
# print(target_category)
self.model.zero_grad()
loss = self.get_loss(output, target_category)
loss.backward(retain_graph=True)
activations = self.activations_and_grads.activations[-1].cpu().data.numpy()[0, :]
grads = self.activations_and_grads.gradients[-1].cpu().data.numpy()[0, :]
#weights = np.mean(grads, axis=(0))
weights = self.get_cam_weights(input_tensor, target_category, activations, grads)
cam = np.zeros(activations.shape[1:], dtype=np.float32)
for i, w in enumerate(weights):
cam += w * activations[i, :]
# cam = activations.T.dot(weights)
# cam = activations.dot(weights)
# cam = activations.dot(weights)
# print(input_tensor.shape[1])
# print(cam.shape)
# x = np.arange(0, 247, 1)
# plt.plot(x, cam.reshape(-1, 1))
# sns.set()
# ax = sns.heatmap(cam.reshape(-1, 1).T)
#cam = cv2.resize(cam, input_tensor.shape[1:][::-1])
#cam = resize_1d(cam, (input_tensor.shape[2]))
cam = np.interp(np.linspace(0, cam.shape[0], input_tensor.shape[2]), np.linspace(0, cam.shape[0], cam.shape[0]), cam) #Change it to the interpolation algorithm that numpy comes with.
#cam = np.maximum(cam, 0)
# cam = np.expand_dims(cam, axis=1)
# ax = sns.heatmap(cam)
# plt.show()
# cam = cam - np.min(cam)
# cam = cam / np.max(cam)
heatmap = (cam - np.min(cam)) / (np.max(cam) - np.min(cam) + 1e-10)#归一化处理
# heatmap = (cam - np.mean(cam, axis=-1)) / (np.std(cam, axis=-1) + 1e-10)
print(heatmap.shape)
return heatmap
class GradCAM(BaseCAM):
def __init__(self, model, target_layer, use_cuda=False):
super(GradCAM, self).__init__(model, target_layer, use_cuda)
def get_cam_weights(self, input_tensor,
target_category,
activations,
grads):
grads_power_2 = grads ** 2
grads_power_3 = grads_power_2 * grads
sum_activations = np.sum(activations, axis=1)
eps = 0.000001
aij = grads_power_2 / (2 * grads_power_2 + sum_activations[:, None] * grads_power_3 + eps)
aij = np.where(grads != 0, aij, 0)
weights = np.maximum(grads, 0) * aij
weights = np.sum(weights, axis=1)
return weights
# from pytorch_grad_cam.utils.image import preprocess_image
import matplotlib.pyplot as plt
def plot_and_save(data1, data2, filename='plot.png'):
"""
Plots two one-dimensional data series on the same plot and saves it as an image.
Parameters:
data1 (list or numpy array): First one-dimensional data series.
data2 (list or numpy array): Second one-dimensional data series.
filename (str): Filename to save the plot image. Default is 'plot.png'.
"""
# Create a new figure
fig, ax1 = plt.subplots()
# Plot data1
ax1.plot(data1, label='Data 1', color='tab:blue')
ax1.set_xlabel('Index')
ax1.set_ylabel('Data 1', color='tab:blue')
ax1.tick_params(axis='y', labelcolor='tab:blue')
# Create a second y-axis for data2
ax2 = ax1.twinx()
ax2.plot(data2, label='Data 2', color='tab:red')
ax2.set_ylabel('Data 2', color='tab:red')
ax2.tick_params(axis='y', labelcolor='tab:red')
# Combine legends
lines, labels = ax1.get_legend_handles_labels()
lines2, labels2 = ax2.get_legend_handles_labels()
ax2.legend(lines + lines2, labels + labels2, loc='upper right')
# Save the plot as an image
plt.savefig(filename)
# Close the plot to release resources
plt.close()
if __name__=="__main__":
import argparse
import os
parser = argparse.ArgumentParser(description='Progressive Growing of GANs')
parser.add_argument('--log_name', type=str, default='EXP6', help='Read and Save log file name')
parser.add_argument('--save_path', type=str, default='./logFile/EXP6/c/test/plots/', help='input data directoy containing .npy files')
parser.add_argument('--data_dir', type=str, default='data/data_warwick/', help='input data directoy containing .npy files')
parser.add_argument('--fold_name', type=str, default='0', help='input data directoy containing .npy files')
parser.add_argument('--weight', type=str, default="./logFile/EXP6/c/incep/last_0.pt", help='input data directoy containing .npy files')
parser.add_argument('--baseline', action='store_true', help='Is it a baseline corected data.')
parser.add_argument('--FC', action='store_true', help='Set the dataset format to (n, l). Where n is number of sample and l is length.')
args = parser.parse_args()
from model import InceptionNetwork, SpectroscopyTransformerEncoder_PreT, InceptionNetwork_PreT, SpectroscopyTransformerEncoder
# from model import SpectroscopyTransformerEncoder_PreT, InceptionNetwork_PreT
model = InceptionNetwork(4)
model.load_state_dict(torch.load(args.weight)['model'])
target_layer = model.inception1
# model = InceptionNetwork_PreT(4)
# model.load_state_dict(torch.load(args.weight)['model'])
# target_layer = model.IR_PreT.inception1
# model = SpectroscopyTransformerEncoder_PreT(num_classes=4, mlp_size=64)
# model.load_state_dict(torch.load(args.weight)['model'])
# target_layer = model.IR_PreT.transformer_encoder
# model = SpectroscopyTransformerEncoder(num_classes=4, mlp_size=64)
# model.load_state_dict(torch.load(args.weight)['model'])
# target_layer = model.transformer_encoder
net = GradCAM(model, target_layer)
# args.baseline = True
# for fold_name in os.listdir(f'logFile/{args.log_name}/data_split'):
# args.fold_num = int(fold_name)
folder_path = os.path.join(f'logFile/{args.log_name}/data_split', args.fold_name)
dataarrayX_val, dataarrayY_val, data_number_val, data_list = dataset_from_txt(args, os.path.join(folder_path, "index_val.txt"))
data_names = data_list
dataset_val = IRDatasetFromNames(dataarrayX_val, dataarrayY_val, data_number_val, FC=args.FC)
val_dataloader = torch.utils.data.DataLoader(dataset_val, batch_size=1, shuffle=False, drop_last=True)
print(f"Validation data size {len(dataset_val)}")
# test(args, model, val_dataloader, weight_path, n_classes=len(np.unique(dataarrayY_val)), arch=args.model)
heatmap0, heatmap1, heatmap2, heatmap3 = None, None, None, None
input_tensors0, input_tensors1, input_tensors2, input_tensors3 = None, None, None, None
for input_tensor, labels, _ in val_dataloader:
# inputs, labels = inputs.to(device), labels.to(device)
# outputs = model(inputs)
# _, predicted = torch.max(outputs.data, 1)
output = net(input_tensor)
input_tensor1 = input_tensor.numpy().squeeze()
if labels==0:
if heatmap0 is None:
heatmap0 = output
input_tensors0 = input_tensor1
else:
heatmap0 = np.vstack([heatmap0, output])
input_tensors0 = np.vstack([input_tensors0, input_tensor1])
if labels==1:
if heatmap1 is None:
heatmap1 = output
input_tensors1 = input_tensor1
else:
heatmap1 = np.vstack([heatmap1, output])
input_tensors1 = np.vstack([input_tensors1, input_tensor1])
if labels==2:
if heatmap2 is None:
heatmap2 = output
input_tensors2 = input_tensor1
else:
heatmap2 = np.vstack([heatmap2, output])
input_tensors2 = np.vstack([input_tensors2, input_tensor1])
if labels==3:
if heatmap3 is None:
heatmap3 = output
input_tensors3 = input_tensor1
else:
heatmap3 = np.vstack([heatmap3, output])
input_tensors3 = np.vstack([input_tensors3, input_tensor1])
if not os.path.exists(args.save_path):
os.mkdir(f"{args.save_path}")
os.mkdir(f"{args.save_path}{args.fold_name}")
np.save(f"{args.save_path}{args.fold_name}/heatmap0.npy", heatmap0)
np.save(f"{args.save_path}{args.fold_name}/heatmap1.npy", heatmap1)
np.save(f"{args.save_path}{args.fold_name}/heatmap2.npy", heatmap2)
np.save(f"{args.save_path}{args.fold_name}/heatmap3.npy", heatmap3)
np.save(f"{args.save_path}{args.fold_name}/input_tensors0.npy", input_tensors0)
np.save(f"{args.save_path}{args.fold_name}/input_tensors1.npy", input_tensors1)
np.save(f"{args.save_path}{args.fold_name}/input_tensors2.npy", input_tensors2)
np.save(f"{args.save_path}{args.fold_name}/input_tensors3.npy", input_tensors3)
for i in range(len(heatmap0)):
plot_and_save(heatmap0[i], input_tensors0[i], f"{args.save_path}{args.fold_name}/0_{i}.png")
for i in range(len(heatmap1)):
plot_and_save(heatmap1[i], input_tensors1[i], f"{args.save_path}{args.fold_name}/1_{i}.png")
for i in range(len(heatmap2)):
plot_and_save(heatmap2[i], input_tensors2[i], f"{args.save_path}{args.fold_name}/2_{i}.png")
for i in range(len(heatmap3)):
plot_and_save(heatmap3[i], input_tensors3[i], f"{args.save_path}{args.fold_name}/3_{i}.png")
plot_and_save(np.mean(heatmap0, axis=0), np.mean(input_tensors0, axis=0), f"{args.save_path}{args.fold_name}/0_mean.png")
plot_and_save(np.mean(heatmap1, axis=0), np.mean(input_tensors1, axis=0), f"{args.save_path}{args.fold_name}/1_mean.png")
plot_and_save(np.mean(heatmap2, axis=0), np.mean(input_tensors2, axis=0), f"{args.save_path}{args.fold_name}/2_mean.png")
plot_and_save(np.mean(heatmap3, axis=0), np.mean(input_tensors3, axis=0), f"{args.save_path}{args.fold_name}/3_mean.png")
# ============================================
# input_tensors = np.load("data/data_war_sngp/C5/PET.npy")
# heatmap = None
# for i in range(input_tensors.shape[0]):
# input_tensor = input_tensors[i,:]
# input_tensor = torch.from_numpy(input_tensor).unsqueeze(dim=0).unsqueeze(dim=0)
# input_tensor = torch.tensor(input_tensor, dtype=torch.float32)
# print(input_tensor.size)
# output = net(input_tensor)
# input_tensor1 = input_tensor.numpy().squeeze()
# print(input_tensor.shape)
# if heatmap is None:
# heatmap = output
# else:
# heatmap = np.vstack([heatmap, output])
# print(heatmap.shape)
# # print(input_tensors.shape)
# plot_and_save(np.mean(heatmap, axis=0), np.mean(input_tensors, axis=0), "logFile/EXP5/cam_PET.png")
# ============================================