forked from lppllppl920/EndoscopyDepthEstimation-Pytorch
-
Notifications
You must be signed in to change notification settings - Fork 0
/
evaluate.py
executable file
·346 lines (314 loc) · 21.2 KB
/
evaluate.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
340
341
342
343
344
345
346
'''
Author: Xingtong Liu, Ayushi Sinha, Masaru Ishii, Gregory D. Hager, Austin Reiter, Russell H. Taylor, and Mathias Unberath
Copyright (C) 2019 Johns Hopkins University - All Rights Reserved
You may use, distribute and modify this code under the
terms of the GNU GENERAL PUBLIC LICENSE Version 3 license for non-commercial usage.
You should have received a copy of the GNU GENERAL PUBLIC LICENSE Version 3 license with
'''
import tqdm
import cv2
import numpy as np
from pathlib import Path
import torchsummary
import torch
import random
from tensorboardX import SummaryWriter
import albumentations as albu
import argparse
import datetime
# Local
import models
import utils
import dataset
if __name__ == '__main__':
cv2.destroyAllWindows()
parser = argparse.ArgumentParser(
description='Self-supervised Depth Estimation on Monocular Endoscopy Dataset -- Evaluate',
formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('--input_downsampling', type=float, default=4.0,
help='image downsampling rate')
parser.add_argument('--input_size', nargs='+', type=int, required=True,
help='input size')
parser.add_argument('--selected_frame_index_list', nargs='+', type=int, required=False, default=None,
help='selected frame index list)')
parser.add_argument('--batch_size', type=int, default=1, help='batch size for testing')
parser.add_argument('--num_workers', type=int, default=2, help='number of workers for input data loader')
parser.add_argument('--num_pre_workers', type=int, default=8,
help='number of workers for preprocessing intermediate data')
parser.add_argument('--adjacent_range', nargs='+', type=int, required=True,
help='interval range for a pair of video frames')
parser.add_argument('--id_range', nargs='+', type=int, required=True,
help='id range for the training and testing dataset')
parser.add_argument('--network_downsampling', type=int, default=64, help='downsampling of network')
parser.add_argument('--inlier_percentage', type=float, default=0.995,
help='percentage of inliers of SfM point clouds (for pruning some outliers)')
parser.add_argument('--testing_patient_id', nargs='+', type=int, help='id of the testing patient')
parser.add_argument('--load_intermediate_data', action='store_true', help='whether to load intermediate data')
parser.add_argument('--use_hsv_colorspace', action='store_true',
help='convert RGB to hsv colorspace')
parser.add_argument('--architecture_summary', action='store_true', help='display the network architecture')
parser.add_argument('--load_all_frames', action='store_true',
help='whether or not to load all frames in sequence root')
parser.add_argument('--trained_model_path', type=str, required=True, help='path to the trained student model')
parser.add_argument('--sequence_root', type=str, required=True, help='path to the testing sequence')
parser.add_argument('--evaluation_result_root', type=str, required=True,
help='logging root')
parser.add_argument('--evaluation_data_root', type=str, required=True, help='path to the testing data')
parser.add_argument('--phase', type=str, required=True, help='phase')
parser.add_argument('--visibility_overlap', type=int, default=30, help='overlap of point visibility information')
args = parser.parse_args()
# Fix randomness for reproducibility
torch.backends.cudnn.deterministic = True
torch.backends.cudnn.benchmark = False
torch.manual_seed(10085)
np.random.seed(10085)
random.seed(10085)
# Hyper-parameters
height, width = args.input_size
adjacent_range = args.adjacent_range
id_range = args.id_range
input_downsampling = args.input_downsampling
batch_size = args.batch_size
num_workers = args.num_workers
network_downsampling = args.network_downsampling
inlier_percentage = args.inlier_percentage
testing_patient_id = args.testing_patient_id
load_intermediate_data = args.load_intermediate_data
is_hsv = args.use_hsv_colorspace
display_architecture = args.architecture_summary
selected_frame_index_list = args.selected_frame_index_list
load_all_frames = args.load_all_frames
phase = args.phase
evaluation_result_root = Path(args.evaluation_result_root)
evaluation_data_root = Path(args.evaluation_data_root)
trained_model_path = Path(args.trained_model_path)
sequence_root = Path(args.sequence_root)
visibility_overlap = args.visibility_overlap
num_pre_workers = args.num_pre_workers
currentDT = datetime.datetime.now()
log_root = Path(evaluation_result_root) / "depth_estimation_evaluation_run_{}_{}_{}_{}_test_id_{}".format(
currentDT.month,
currentDT.day,
currentDT.hour,
currentDT.minute,
"_".join(testing_patient_id))
if not log_root.exists():
log_root.mkdir(parents=True)
writer = SummaryWriter(logdir=str(log_root))
print("Tensorboard visualization at {}".format(str(log_root)))
if selected_frame_index_list is None and not load_all_frames:
raise IOError
# Read all frame indexes
if load_all_frames:
selected_frame_index_list = utils.read_visible_view_indexes(sequence_root)
# Get color image filenames
test_filenames = utils.get_filenames_from_frame_indexes(sequence_root, selected_frame_index_list)
folder_list = utils.get_parent_folder_names(evaluation_data_root, id_range=id_range)
if phase == "validation":
test_dataset = dataset.SfMDataset(image_file_names=test_filenames,
folder_list=folder_list,
adjacent_range=adjacent_range, transform=None,
downsampling=input_downsampling,
network_downsampling=network_downsampling,
inlier_percentage=inlier_percentage,
use_store_data=load_intermediate_data,
store_data_root=evaluation_data_root,
phase="validation", is_hsv=is_hsv,
num_pre_workers=num_pre_workers, visible_interval=visibility_overlap,
rgb_mode="rgb")
test_loader = torch.utils.data.DataLoader(dataset=test_dataset, batch_size=batch_size, shuffle=False,
num_workers=0)
depth_estimation_model = models.FCDenseNet57(n_classes=1)
# Initialize the depth estimation network with Kaiming He initialization
utils.init_net(depth_estimation_model, type="kaiming", mode="fan_in", activation_mode="relu",
distribution="normal")
# Multi-GPU running
depth_estimation_model = torch.nn.DataParallel(depth_estimation_model)
# Summary network architecture
if display_architecture:
torchsummary.summary(depth_estimation_model, input_size=(3, height, width))
# Load trained model
if trained_model_path.exists():
print("Loading {:s} ...".format(str(trained_model_path)))
state = torch.load(str(trained_model_path))
step = state['step']
epoch = state['epoch']
depth_estimation_model.load_state_dict(state['model'])
print('Restored model, epoch {}, step {}'.format(epoch, step))
else:
print("Trained model could not be found")
raise OSError
depth_estimation_model = depth_estimation_model.module
# Custom layers
depth_scaling_layer = models.DepthScalingLayer()
depth_warping_layer = models.DepthWarpingLayer()
flow_from_depth_layer = models.FlowfromDepthLayer()
with torch.no_grad():
# Set model to evaluation mode
depth_estimation_model.eval()
# Update progress bar
tq = tqdm.tqdm(total=len(test_loader) * batch_size)
for batch, (
colors_1, colors_2, sparse_depths_1, sparse_depths_2, sparse_depth_masks_1, sparse_depth_masks_2,
sparse_flows_1, sparse_flows_2, sparse_flow_masks_1, sparse_flow_masks_2, boundaries,
rotations_1_wrt_2, rotations_2_wrt_1, translations_1_wrt_2, translations_2_wrt_1, intrinsics,
folders) in enumerate(test_loader):
colors_1 = colors_1.cuda()
colors_2 = colors_2.cuda()
sparse_depths_1 = sparse_depths_1.cuda()
sparse_depths_2 = sparse_depths_2.cuda()
sparse_depth_masks_1 = sparse_depth_masks_1.cuda()
sparse_depth_masks_2 = sparse_depth_masks_2.cuda()
sparse_flows_1 = sparse_flows_1.cuda()
sparse_flows_2 = sparse_flows_2.cuda()
boundaries = boundaries.cuda()
rotations_1_wrt_2 = rotations_1_wrt_2.cuda()
rotations_2_wrt_1 = rotations_2_wrt_1.cuda()
translations_1_wrt_2 = translations_1_wrt_2.cuda()
translations_2_wrt_1 = translations_2_wrt_1.cuda()
intrinsics = intrinsics.cuda()
tq.update(batch_size)
colors_1 = boundaries * colors_1
colors_2 = boundaries * colors_2
sparse_flows_1 = sparse_flows_1 * boundaries
sparse_flows_2 = sparse_flows_2 * boundaries
predicted_depth_maps_1 = depth_estimation_model(colors_1)
predicted_depth_maps_2 = depth_estimation_model(colors_2)
scaled_depth_maps_1, normalized_scale_std_1 = depth_scaling_layer(
[torch.abs(predicted_depth_maps_1), sparse_depths_1, sparse_depth_masks_1])
scaled_depth_maps_2, normalized_scale_std_2 = depth_scaling_layer(
[torch.abs(predicted_depth_maps_2), sparse_depths_2, sparse_depth_masks_2])
depth_array = scaled_depth_maps_1[0].squeeze(dim=0).data.cpu().numpy()
color_array = np.uint8(255 * cv2.cvtColor(
(colors_1[0].permute(1, 2, 0).data.cpu().numpy() + 1.0) * 0.5, cv2.COLOR_HSV2BGR_FULL))
boundary_array = boundaries[0].squeeze(dim=0).data.cpu().numpy()
intrinsic_array = intrinsics[0].data.cpu().numpy()
# Sparse flow loss
flows_from_depth_1 = flow_from_depth_layer(
[scaled_depth_maps_1, boundaries, translations_1_wrt_2, rotations_1_wrt_2,
intrinsics])
flows_from_depth_2 = flow_from_depth_layer(
[scaled_depth_maps_2, boundaries, translations_2_wrt_1, rotations_2_wrt_1,
intrinsics])
flows_from_depth_1 = flows_from_depth_1 * boundaries
flows_from_depth_2 = flows_from_depth_2 * boundaries
warped_depth_maps_2_to_1, intersect_masks_1 = depth_warping_layer(
[scaled_depth_maps_1, scaled_depth_maps_2, boundaries, translations_1_wrt_2, rotations_1_wrt_2,
intrinsics])
warped_depth_maps_1_to_2, intersect_masks_2 = depth_warping_layer(
[scaled_depth_maps_2, scaled_depth_maps_1, boundaries, translations_2_wrt_1, rotations_2_wrt_1,
intrinsics])
colors_1_display, sparse_depths_1_display, pred_depths_1_display, warped_depths_1_display, sparse_flows_1_display, dense_flows_1_display = \
utils.display_color_sparse_depth_dense_depth_warped_depth_sparse_flow_dense_flow(idx=1, step=step,
writer=writer,
colors_1=colors_1,
sparse_depths_1=sparse_depths_1,
pred_depths_1=scaled_depth_maps_1 * boundaries,
warped_depths_2_to_1=warped_depth_maps_2_to_1,
sparse_flows_1=sparse_flows_1,
flows_from_depth_1=flows_from_depth_1,
phase="validation",
is_return_image=True,
color_reverse=True,
is_hsv=is_hsv,
rgb_mode="rgb",
boundaries=boundaries
)
colors_2_display, sparse_depths_2_display, pred_depths_2_display, warped_depths_2_display, sparse_flows_2_display, dense_flows_2_display = \
utils.display_color_sparse_depth_dense_depth_warped_depth_sparse_flow_dense_flow(idx=2, step=step,
writer=writer,
colors_1=colors_2,
sparse_depths_1=sparse_depths_2,
pred_depths_1=scaled_depth_maps_2 * boundaries,
warped_depths_2_to_1=warped_depth_maps_1_to_2,
sparse_flows_1=sparse_flows_2,
flows_from_depth_1=flows_from_depth_2,
phase="validation",
is_return_image=True,
color_reverse=True,
is_hsv=is_hsv,
rgb_mode="rgb",
boundaries=boundaries
)
image_display = utils.stack_and_display(phase="validation",
title="Results (c1, sd1, d1, wd1, sf1, df1, c2, sd2, d2, wd2, sf2, df2)",
step=step, writer=writer,
image_list=[colors_1_display, sparse_depths_1_display,
pred_depths_1_display,
warped_depths_1_display, sparse_flows_1_display,
dense_flows_1_display,
colors_2_display, sparse_depths_2_display,
pred_depths_2_display,
warped_depths_2_display, sparse_flows_2_display,
dense_flows_2_display],
return_image=True)
cv2.imwrite(str(log_root / "{}.png".format(batch)),
cv2.cvtColor(np.uint8(image_display * 255), cv2.COLOR_RGB2BGR))
point_cloud = utils.point_cloud_from_depth(depth_array, color_array, boundary_array,
intrinsic_array, point_cloud_downsampling=1)
utils.write_point_cloud(str(log_root / "{}.ply".format(batch)), point_cloud)
tq.close()
writer.close()
elif phase == "test":
test_dataset = dataset.SfMDataset(image_file_names=test_filenames,
folder_list=folder_list,
adjacent_range=adjacent_range, transform=None,
downsampling=input_downsampling,
network_downsampling=network_downsampling,
inlier_percentage=inlier_percentage,
use_store_data=load_intermediate_data,
store_data_root=evaluation_data_root,
phase="test", is_hsv=is_hsv,
num_pre_workers=num_pre_workers, visible_interval=visibility_overlap,
rgb_mode="rgb")
test_loader = torch.utils.data.DataLoader(dataset=test_dataset, batch_size=1, shuffle=False,
num_workers=0)
depth_estimation_model = models.FCDenseNet57(n_classes=1)
# Initialize the depth estimation network with Kaiming He initialization
utils.init_net(depth_estimation_model, type="kaiming", mode="fan_in", activation_mode="relu",
distribution="normal")
# Multi-GPU running
depth_estimation_model = torch.nn.DataParallel(depth_estimation_model)
# Summary network architecture
if display_architecture:
torchsummary.summary(depth_estimation_model, input_size=(3, height, width))
# Load trained model
if trained_model_path.exists():
print("Loading {:s} ...".format(str(trained_model_path)))
state = torch.load(str(trained_model_path))
step = state['step']
epoch = state['epoch']
depth_estimation_model.load_state_dict(state['model'])
print('Restored model, epoch {}, step {}'.format(epoch, step))
else:
print("Trained model does not exist")
raise OSError
depth_estimation_model = depth_estimation_model.module
with torch.no_grad():
# Set model to evaluation mode
depth_estimation_model.eval()
# Update progress bar
tq = tqdm.tqdm(total=len(test_loader) * batch_size)
for batch, (colors_1, boundaries, intrinsics, names) in enumerate(test_loader):
colors_1 = colors_1.cuda()
boundaries = boundaries.cuda()
colors_1 = boundaries * colors_1
predicted_depth_maps_1 = depth_estimation_model(colors_1)
color_display = np.uint8(
255 * (0.5 * colors_1[0].permute(1, 2, 0).data.cpu().numpy() + 0.5).reshape((height, width, 3)))
if is_hsv:
color_display = cv2.cvtColor(color_display, cv2.COLOR_HSV2BGR_FULL)
else:
color_display = cv2.cvtColor(color_display, cv2.COLOR_RGB2BGR)
boundary = boundaries[0].data.cpu().numpy().reshape((height, width))
color_display = np.uint8(boundary.reshape((height, width, 1)) * color_display)
depth_map = (boundaries * predicted_depth_maps_1)[0].data.cpu().numpy().reshape((height, width))
depth_display = cv2.applyColorMap(np.uint8(255 * depth_map / np.max(depth_map)), cv2.COLORMAP_JET)
point_cloud = utils.point_cloud_from_depth(depth_map=depth_map, color_img=color_display,
mask_img=boundary,
intrinsic_matrix=intrinsics[0].data.numpy(),
point_cloud_downsampling=1)
utils.write_point_cloud(path=log_root / "{}.ply".format(names[0]), point_cloud=point_cloud)
cv2.imwrite(str(log_root / "{}.png".format(names[0])), cv2.hconcat([color_display, depth_display]))
tq.update(batch_size)