-
Notifications
You must be signed in to change notification settings - Fork 86
/
hipt_4k.py
330 lines (274 loc) · 15.4 KB
/
hipt_4k.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
### Dependencies
# Base Dependencies
import os
import pickle
import sys
# LinAlg / Stats / Plotting Dependencies
import h5py
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
from PIL import Image
Image.MAX_IMAGE_PIXELS = None
from tqdm import tqdm
# Torch Dependencies
import torch
import torch.multiprocessing
import torchvision
from torchvision import transforms
from einops import rearrange, repeat
torch.multiprocessing.set_sharing_strategy('file_system')
# Local Dependencies
import vision_transformer as vits
import vision_transformer4k as vits4k
from hipt_heatmap_utils import *
from hipt_model_utils import get_vit256, get_vit4k, tensorbatch2im, eval_transforms, roll_batch2img
class HIPT_4K(torch.nn.Module):
"""
HIPT Model (ViT-4K) for encoding non-square images (with [256 x 256] patch tokens), with
[256 x 256] patch tokens encoded via ViT-256 using [16 x 16] patch tokens.
"""
def __init__(self,
model256_path: str = '../Checkpoints/vit256_small_dino.pth',
model4k_path: str = '../Checkpoints/vit4k_xs_dino.pth',
device256=torch.device('cuda:0'),
device4k=torch.device('cuda:1')):
super().__init__()
self.model256 = get_vit256(pretrained_weights=model256_path).to(device256)
self.model4k = get_vit4k(pretrained_weights=model4k_path).to(device4k)
self.device256 = device256
self.device4k = device4k
def forward(self, x):
"""
Forward pass of HIPT (given an image tensor x), outputting the [CLS] token from ViT-4K.
1. x is center-cropped such that the W / H is divisible by the patch token size in ViT-4K (e.g. - 256 x 256).
2. x then gets unfolded into a "batch" of [256 x 256] images.
3. A pretrained ViT-256 model extracts the CLS token from each [256 x 256] image in the batch.
4. These batch-of-features are then reshaped into a 2D feature grid (of width "w_256" and height "h_256".)
5. This feature grid is then used as the input to ViT-4K, outputting [CLS]_4K.
Args:
- x (torch.Tensor): [1 x C x W' x H'] image tensor.
Return:
- features_cls4k (torch.Tensor): [1 x 192] cls token (d_4k = 192 by default).
"""
batch_256, w_256, h_256 = self.prepare_img_tensor(x) # 1. [1 x 3 x W x H]
batch_256 = batch_256.unfold(2, 256, 256).unfold(3, 256, 256) # 2. [1 x 3 x w_256 x h_256 x 256 x 256]
batch_256 = rearrange(batch_256, 'b c p1 p2 w h -> (b p1 p2) c w h') # 2. [B x 3 x 256 x 256], where B = (1*w_256*h_256)
features_cls256 = []
for mini_bs in range(0, batch_256.shape[0], 256): # 3. B may be too large for ViT-256. We further take minibatches of 256.
minibatch_256 = batch_256[mini_bs:mini_bs+256].to(self.device256, non_blocking=True)
features_cls256.append(self.model256(minibatch_256).detach().cpu()) # 3. Extracting ViT-256 features from [256 x 3 x 256 x 256] image batches.
features_cls256 = torch.vstack(features_cls256) # 3. [B x 384], where 384 == dim of ViT-256 [ClS] token.
features_cls256 = features_cls256.reshape(w_256, h_256, 384).transpose(0,1).transpose(0,2).unsqueeze(dim=0)
features_cls256 = features_cls256.to(self.device4k, non_blocking=True) # 4. [1 x 384 x w_256 x h_256]
features_cls4k = self.model4k.forward(features_cls256) # 5. [1 x 192], where 192 == dim of ViT-4K [ClS] token.
return features_cls4k
def forward_asset_dict(self, x: torch.Tensor):
"""
Forward pass of HIPT (given an image tensor x), with certain intermediate representations saved in
a dictionary (that is to be stored in a H5 file). See walkthrough of how the model works above.
Args:
- x (torch.Tensor): [1 x C x W' x H'] image tensor.
Return:
- asset_dict (dict): Dictionary of intermediate feature representations of HIPT and other metadata.
- features_cls256 (np.array): [B x 384] extracted ViT-256 cls tokens
- features_mean256 (np.array): [1 x 384] mean ViT-256 cls token (exluding non-tissue patches)
- features_4k (np.array): [1 x 192] extracted ViT-4K cls token.
- features_4k (np.array): [1 x 576] feature vector (concatenating mean ViT-256 + ViT-4K cls tokens)
"""
batch_256, w_256, h_256 = self.prepare_img_tensor(x)
batch_256 = batch_256.unfold(2, 256, 256).unfold(3, 256, 256)
batch_256 = rearrange(batch_256, 'b c p1 p2 w h -> (b p1 p2) c w h')
features_cls256 = []
for mini_bs in range(0, batch_256.shape[0], 256):
minibatch_256 = batch_256[mini_bs:mini_bs+256].to(self.device256, non_blocking=True)
features_cls256.append(self.model256(minibatch_256).detach().cpu())
features_cls256 = torch.vstack(features_cls256)
features_mean256 = features_cls256.mean(dim=0).unsqueeze(dim=0)
features_grid256 = features_cls256.reshape(w_256, h_256, 384).transpose(0,1).transpose(0,2).unsqueeze(dim=0)
features_grid256 = features_grid256.to(self.device4k, non_blocking=True)
features_cls4k = self.model4k.forward(features_grid256).detach().cpu()
features_mean256_cls4k = torch.cat([features_mean256, features_cls4k], dim=1)
asset_dict = {
'features_cls256': features_cls256.numpy(),
'features_mean256': features_mean256.numpy(),
'features_cls4k': features_cls4k.numpy(),
'features_mean256_cls4k': features_mean256_cls4k.numpy()
}
return asset_dict
def _get_region_attention_scores(self, region, scale=1):
r"""
Forward pass in hierarchical model with attention scores saved.
Args:
- region (PIL.Image): 4096 x 4096 Image
- model256 (torch.nn): 256-Level ViT
- model4k (torch.nn): 4096-Level ViT
- scale (int): How much to scale the output image by (e.g. - scale=4 will resize images to be 1024 x 1024.)
Returns:
- np.array: [256, 256/scale, 256/scale, 3] np.array sequence of image patches from the 4K x 4K region.
- attention_256 (torch.Tensor): [256, 256/scale, 256/scale, 3] torch.Tensor sequence of attention maps for 256-sized patches.
- attention_4k (torch.Tensor): [1, 4096/scale, 4096/scale, 3] torch.Tensor sequence of attention maps for 4k-sized regions.
"""
eval_t = transforms.Compose([transforms.ToTensor(), transforms.Normalize([0.5, 0.5, 0.5], [0.5, 0.5, 0.5])])
x = eval_transforms()(region).unsqueeze(dim=0)
batch_256, w_256, h_256 = self.prepare_img_tensor(x)
batch_256 = batch_256.unfold(2, 256, 256).unfold(3, 256, 256)
batch_256 = rearrange(batch_256, 'b c p1 p2 w h -> (b p1 p2) c w h')
batch_256 = batch_256.to(self.device256, non_blocking=True)
features_cls256 = self.model256(batch_256)
attention_256 = self.model256.get_last_selfattention(batch_256)
nh = attention_256.shape[1] # number of head
attention_256 = attention_256[:, :, 0, 1:].reshape(256, nh, -1)
attention_256 = attention_256.reshape(w_256*h_256, nh, 16, 16)
attention_256 = nn.functional.interpolate(attention_256, scale_factor=int(16/scale), mode="nearest").cpu().numpy()
features_grid256 = features_cls256.reshape(w_256, h_256, 384).transpose(0,1).transpose(0,2).unsqueeze(dim=0)
features_grid256 = features_grid256.to(self.device4k, non_blocking=True)
features_cls4k = self.model4k.forward(features_grid256).detach().cpu()
attention_4k = self.model4k.get_last_selfattention(features_grid256)
nh = attention_4k.shape[1] # number of head
attention_4k = attention_4k[0, :, 0, 1:].reshape(nh, -1)
attention_4k = attention_4k.reshape(nh, w_256, h_256)
attention_4k = nn.functional.interpolate(attention_4k.unsqueeze(0), scale_factor=int(256/scale), mode="nearest")[0].cpu().numpy()
if scale != 1:
batch_256 = nn.functional.interpolate(batch_256, scale_factor=(1/scale), mode="nearest")
return tensorbatch2im(batch_256), attention_256, attention_4k
def get_region_attention_heatmaps(self, x, offset=128, scale=4, alpha=0.5, cmap = cmap_map(lambda x: x/2 + 0.5, matplotlib.cm.jet), threshold=None):
r"""
Creates hierarchical heatmaps (Raw H&E + ViT-256 + ViT-4K + Blended Heatmaps saved individually).
Args:
- region (PIL.Image): 4096 x 4096 Image
- model256 (torch.nn): 256-Level ViT
- model4k (torch.nn): 4096-Level ViT
- output_dir (str): Save directory / subdirectory
- fname (str): Naming structure of files
- offset (int): How much to offset (from top-left corner with zero-padding) the region by for blending
- scale (int): How much to scale the output image by
- alpha (float): Image blending factor for cv2.addWeighted
- cmap (matplotlib.pyplot): Colormap for creating heatmaps
Returns:
- None
"""
region = Image.fromarray(tensorbatch2im(x)[0])
w, h = region.size
region2 = add_margin(region.crop((128,128,w,h)),
top=0, left=0, bottom=128, right=128, color=(255,255,255))
region3 = add_margin(region.crop((128*2,128*2,w,h)),
top=0, left=0, bottom=128*2, right=128*2, color=(255,255,255))
region4 = add_margin(region.crop((128*3,128*3,w,h)),
top=0, left=0, bottom=128*4, right=128*4, color=(255,255,255))
b256_1, a256_1, a4k_1 = self._get_region_attention_scores(region, scale)
b256_2, a256_2, a4k_2 = self._get_region_attention_scores(region, scale)
b256_3, a256_3, a4k_3 = self._get_region_attention_scores(region, scale)
b256_4, a256_4, a4k_4 = self._get_region_attention_scores(region, scale)
offset_2 = (offset*1)//scale
offset_3 = (offset*2)//scale
offset_4 = (offset*3)//scale
w_s, h_s = w//scale, h//scale
w_256, h_256 = w//256, h//256
save_region = np.array(region.resize((w_s, h_s)))
if threshold != None:
for i in range(6):
score256_1 = concat_scores256(a256_1[:,i,:,:], w_256, h_256, size=(w_s//w_256,h_s//h_256))
score256_2 = concat_scores256(a256_2[:,i,:,:], w_256, h_256, size=(w_s//w_256,h_s//h_256))
new_score256_2 = np.zeros_like(score256_2)
new_score256_2[offset_2:w_s, offset_2:h_s] = score256_2[:(w_s-offset_2), :(h_s-offset_2)]
overlay256 = np.ones_like(score256_2)*100
overlay256[offset_2:w_s, offset_2:h_s] += 100
score256 = (score256_1+new_score256_2)/overlay256
mask256 = score256.copy()
mask256[mask256 < threshold] = 0
mask256[mask256 > threshold] = 0.95
color_block256 = (cmap(mask256)*255)[:,:,:3].astype(np.uint8)
region256_hm = cv2.addWeighted(color_block256, alpha, save_region.copy(), 1-alpha, 0, save_region.copy())
region256_hm[mask256==0] = 0
img_inverse = save_region.copy()
img_inverse[mask256 == 0.95] = 0
Image.fromarray(region256_hm+img_inverse).save(os.path.join(output_dir, '%s_256th[%d].png' % (fname, i)))
if False:
for j in range(6):
score4k_1 = concat_scores4k(a4k_1[j], size=(h_s,w_s))
score4k = score4k_1 / 100
color_block4k = (cmap(score4k)*255)[:,:,:3].astype(np.uint8)
region4k_hm = cv2.addWeighted(color_block4k, alpha, save_region.copy(), 1-alpha, 0, save_region.copy())
Image.fromarray(region4k_hm).save(os.path.join(output_dir, '%s_4k[%s].png' % (fname, j)))
hm4k, hm256, hm4k_256 = [], [], []
for j in range(6):
score4k_1 = concat_scores4k(a4k_1[j], size=(h_s,w_s))
score4k_2 = concat_scores4k(a4k_2[j], size=(h_s,w_s))
score4k_3 = concat_scores4k(a4k_3[j], size=(h_s,w_s))
score4k_4 = concat_scores4k(a4k_4[j], size=(h_s,w_s))
new_score4k_2 = np.zeros_like(score4k_2)
new_score4k_2[offset_2:h_s, offset_2:w_s] = score4k_2[:(h_s-offset_2), :(w_s-offset_2)]
new_score4k_3 = np.zeros_like(score4k_3)
new_score4k_3[offset_3:h_s, offset_3:w_s] = score4k_3[:(h_s-offset_3), :(w_s-offset_3)]
new_score4k_4 = np.zeros_like(score4k_4)
new_score4k_4[offset_4:h_s, offset_4:w_s] = score4k_4[:(h_s-offset_4), :(w_s-offset_4)]
overlay4k = np.ones_like(score4k_2)*100
overlay4k[offset_2:h_s, offset_2:w_s] += 100
overlay4k[offset_3:h_s, offset_3:w_s] += 100
overlay4k[offset_4:h_s, offset_4:w_s] += 100
score4k = (score4k_1+new_score4k_2+new_score4k_3+new_score4k_4)/overlay4k
color_block4k = (cmap(score4k)*255)[:,:,:3].astype(np.uint8)
region4k_hm = cv2.addWeighted(color_block4k, alpha, save_region.copy(), 1-alpha, 0, save_region.copy())
hm4k.append(Image.fromarray(region4k_hm))
for i in range(6):
score256_1 = concat_scores256(a256_1[:,i,:,:], h_256, w_256, size=(256, 256))
score256_2 = concat_scores256(a256_2[:,i,:,:], h_256, w_256, size=(256, 256))
new_score256_2 = np.zeros_like(score256_2)
new_score256_2[offset_2:h_s, offset_2:w_s] = score256_2[:(h_s-offset_2), :(w_s-offset_2)]
overlay256 = np.ones_like(score256_2)*100
overlay256[offset_2:h_s, offset_2:w_s] += 100
score256 = (score256_1+new_score256_2)/overlay256
color_block256 = (cmap(score256)*255)[:,:,:3].astype(np.uint8)
region256_hm = cv2.addWeighted(color_block256, alpha, save_region.copy(), 1-alpha, 0, save_region.copy())
hm256.append(Image.fromarray(region256_hm))
for j in range(6):
score4k_1 = concat_scores4k(a4k_1[j], size=(h_s,w_s))
score4k_2 = concat_scores4k(a4k_2[j], size=(h_s,w_s))
score4k_3 = concat_scores4k(a4k_3[j], size=(h_s,w_s))
score4k_4 = concat_scores4k(a4k_4[j], size=(h_s,w_s))
new_score4k_2 = np.zeros_like(score4k_2)
new_score4k_2[offset_2:h_s, offset_2:w_s] = score4k_2[:(h_s-offset_2), :(w_s-offset_2)]
new_score4k_3 = np.zeros_like(score4k_3)
new_score4k_3[offset_3:h_s, offset_3:w_s] = score4k_3[:(h_s-offset_3), :(w_s-offset_3)]
new_score4k_4 = np.zeros_like(score4k_4)
new_score4k_4[offset_4:h_s, offset_4:w_s] = score4k_4[:(h_s-offset_4), :(w_s-offset_4)]
overlay4k = np.ones_like(score4k_2)*100
overlay4k[offset_2:h_s, offset_2:w_s] += 100
overlay4k[offset_3:h_s, offset_3:w_s] += 100
overlay4k[offset_4:h_s, offset_4:w_s] += 100
score4k = (score4k_1+new_score4k_2+new_score4k_3+new_score4k_4)/overlay4k
for i in range(6):
score256_1 = concat_scores256(a256_1[:,i,:,:], h_256, w_256, size=(256, 256))
score256_2 = concat_scores256(a256_2[:,i,:,:], h_256, w_256, size=(256, 256))
new_score256_2 = np.zeros_like(score256_2)
new_score256_2[offset_2:h_s, offset_2:w_s] = score256_2[:(h_s-offset_2), :(w_s-offset_2)]
overlay256 = np.ones_like(score256_2)*100
overlay256[offset_2:h_s, offset_2:w_s] += 100
score256 = (score256_1+new_score256_2)/overlay256
factorize = lambda data: (data - np.min(data)) / (np.max(data) - np.min(data))
score = (score4k*overlay4k+score256*overlay256)/(overlay4k+overlay256) #factorize(score256*score4k)
color_block = (cmap(score)*255)[:,:,:3].astype(np.uint8)
region4k_256_hm = cv2.addWeighted(color_block, alpha, save_region.copy(), 1-alpha, 0, save_region.copy())
hm4k_256.append(Image.fromarray(region4k_256_hm))
return hm4k, hm256, hm4k_256
def prepare_img_tensor(self, img: torch.Tensor, patch_size=256):
"""
Helper function that takes a non-square image tensor, and takes a center crop s.t. the width / height
are divisible by 256.
(Note: "_256" for w / h is should technically be renamed as "_ps", but may not be easier to read.
Until I need to make HIPT with patch_sizes != 256, keeping the naming convention as-is.)
Args:
- img (torch.Tensor): [1 x C x W' x H'] image tensor.
- patch_size (int): Desired patch size to evenly subdivide the image.
Return:
- img_new (torch.Tensor): [1 x C x W x H] image tensor, where W and H are divisble by patch_size.
- w_256 (int): # of [256 x 256] patches of img_new's width (e.g. - W/256)
- h_256 (int): # of [256 x 256] patches of img_new's height (e.g. - H/256)
"""
make_divisble = lambda l, patch_size: (l - (l % patch_size))
b, c, w, h = img.shape
load_size = make_divisble(w, patch_size), make_divisble(h, patch_size)
w_256, h_256 = w // patch_size, h // patch_size
img_new = transforms.CenterCrop(load_size)(img)
return img_new, w_256, h_256