-
Notifications
You must be signed in to change notification settings - Fork 9
Expand file tree
/
Copy pathtrain.py
More file actions
1091 lines (862 loc) · 48.4 KB
/
Copy pathtrain.py
File metadata and controls
1091 lines (862 loc) · 48.4 KB
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
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
#
# Copyright (C) 2023, Inria
# GRAPHDECO research group, https://team.inria.fr/graphdeco
# All rights reserved.
#
# This software is free for non-commercial, research and evaluation use
# under the terms of the LICENSE.md file.
#
# For inquiries contact george.drettakis@inria.fr
#
import os
import numpy as np
import torch
import torchvision
import json
import wandb
import time
from os import makedirs
import shutil, pathlib
from pathlib import Path
from PIL import Image
import torchvision.transforms.functional as tf
import lpips
from random import randint
from utils.loss_utils import l1_loss, ssim, ScaleAndShiftLoss, l1_loss_masked, ssim_masked, l2_loss, my_ssim, binary_cross_entropy_loss
from gaussian_renderer import prefilter_voxel, render, network_gui, prefilter_position2D
import sys
from scene import Scene, GaussianModel
from utils.general_utils import safe_state
import uuid
from tqdm import tqdm
from utils.image_utils import psnr
from argparse import ArgumentParser, Namespace
from arguments import ModelParams, PipelineParams, OptimizationParams
import matplotlib.pyplot as plt
# from utils.loss_utils import compute_scale_and_shift
from utils.colmap_utils import read_cameras_binary
import cv2
import torch.nn.functional as F
import random
import time
# torch.set_num_threads(32)
lpips_fn = lpips.LPIPS(net='vgg').to('cuda')
loss_fn_vgg_spatial = lpips.LPIPS(net='vgg', spatial=True).to('cuda')
def cal_box(mask):
x_coords, y_coords = np.nonzero(mask[:, :]).split([1, 1], dim=1)
x_coords = torch.squeeze(x_coords)
y_coords = torch.squeeze(y_coords)
x_min, y_min = torch.min(x_coords), torch.min(y_coords)
x_max, y_max = torch.max(x_coords), torch.max(y_coords)
return np.asarray([x_min.cpu(), y_min.cpu(), x_max.cpu(), y_max.cpu()])
def expand_bbox_by_ratio(bbox_ori, enlarge_ratio, max_h, max_w, int_=False):
x_min, y_min, x_max, y_max = bbox_ori[0], bbox_ori[1], bbox_ori[2], bbox_ori[3]
# calculate ori bbox's w h
width = x_max - x_min
height = y_max - y_min
# calculate ori bbox's center coor
center_x = (x_max + x_min) / 2
center_y = (y_max + y_min) / 2
# calculate new bbox's w h
new_width = width * enlarge_ratio
new_height = height * enlarge_ratio
# calculate new bbox's x_min, y_min, x_max, y_max
new_x_min = center_x - new_width / 2
new_y_min = center_y - new_height / 2
new_x_max = center_x + new_width / 2
new_y_max = center_y + new_height / 2
if not int_:
new_x_min = max(0, new_x_min)
new_y_min = max(0, new_y_min)
new_x_max = min(new_x_max, max_h-1)
new_y_max = min(new_y_max, max_w-1)
# return coor of new bbox
return new_x_min, new_y_min, new_x_max, new_y_max, width, height, new_width, new_height
else:
new_x_min = max(0, int(new_x_min))
new_y_min = max(0, int(new_y_min))
new_x_max = min(int(new_x_max), max_h-1)
new_y_max = min(int(new_y_max), max_w-1)
# return coor of new bbox
return new_x_min, new_y_min, new_x_max, new_y_max, width, height, new_width, new_height
def get_random_mask(gt_mask, enlarge_ratio, small_ratio, max_h, max_w):
bbox_ori = cal_box(gt_mask) # 0 0 h w
new_x_min, new_y_min, new_x_max, new_y_max, ori_width, ori_height, new_width, new_height = expand_bbox_by_ratio(bbox_ori, enlarge_ratio=enlarge_ratio, max_h=max_h, max_w=max_w) # new_x_min, new_y_min, new_x_max, new_y_max
small_width, small_height = small_ratio * ori_width, small_ratio * ori_height
# calculate bbox's max width & height
max_small_width = new_width - small_width
max_small_height = new_height - small_height
# generate random offset
offset_x = random.uniform(0, max_small_width)
offset_y = random.uniform(0, max_small_height)
# calculate bbox's new x_min, y_min, x_max, y_max
new_x_min = max(0, int(new_x_min + offset_x))
new_y_min = max(0, int(new_y_min + offset_y))
new_x_max = min(int(new_x_min + small_width), max_h-1)
new_y_max = min(int(new_y_min + small_height), max_w-1)
fg_mask = torch.zeros_like(gt_mask).unsqueeze(0)
fg_mask[:, new_x_min:new_x_max, new_y_min:new_y_max] = 1.0
return fg_mask
# sampling for whole image
def sample_patch_in_whole_image(H, W, patch_size=256):
if H < patch_size or W < patch_size:
raise ValueError("Input dimensions should be larger than the patch size.")
x_min = torch.randint(0, W - patch_size + 1, (1,))
y_min = torch.randint(0, H - patch_size + 1, (1,))
x_max = x_min + patch_size
y_max = y_min + patch_size
# return x_min.item(), y_min.item(), x_max.item(), y_max.item()
return y_min.item(), y_max.item(), x_min.item(), x_max.item()
def expand_bbox_256_bysize(bbox_ori, patch_size, max_h, max_w):
x_min, y_min, x_max, y_max = bbox_ori[0], bbox_ori[1], bbox_ori[2], bbox_ori[3]
# calculate ori bbox's w h
width = x_max - x_min
height = y_max - y_min
# calculate ori bbox's center coor
center_x = (x_max + x_min) / 2
center_y = (y_max + y_min) / 2
# calculate new bbox's w h
new_width = patch_size
new_height = patch_size
# calculate new bbox's x_min, y_min, x_max, y_max
new_x_min = center_x - new_width / 2
new_y_min = center_y - new_height / 2
new_x_max = center_x + new_width / 2
new_y_max = center_y + new_height / 2
new_x_min = max(0, new_x_min)
new_y_min = max(0, new_y_min)
new_x_max = min(new_x_max, max_h-1)
new_y_max = min(new_y_max, max_w-1)
# return coor of new bbox
return new_x_min, new_y_min, new_x_max, new_y_max, width, height, new_width, new_height
def sample_patch_in_mask_region(gt_mask, patch_size, small_ratio, max_h, max_w):
bbox_ori = cal_box(gt_mask) # 0 0 h w
new_x_min, new_y_min, new_x_max, new_y_max, ori_width, ori_height, new_width, new_height = expand_bbox_256_bysize(bbox_ori, patch_size=256, max_h=max_h, max_w=max_w) # new_x_min, new_y_min, new_x_max, new_y_max
small_width, small_height = small_ratio * ori_width, small_ratio * ori_height
# calculate bbox's max width & height
max_small_width = new_width - small_width
max_small_height = new_height - small_height
# generate random offset
offset_x = random.uniform(0, max_small_width)
offset_y = random.uniform(0, max_small_height)
# calculate bbox's new x_min, y_min, x_max, y_max
new_x_min = max(0, int(new_x_min + offset_x))
new_y_min = max(0, int(new_y_min + offset_y))
new_x_max = min(int(new_x_min + patch_size), max_h-1)
new_y_max = min(int(new_y_min + patch_size), max_w-1)
return new_x_min, new_x_max, new_y_min, new_y_max
# # copy from MiDaS
def compute_scale_and_shift(prediction, target, mask):
# system matrix: A = [[a_00, a_01], [a_10, a_11]]
a_00 = torch.sum(mask * prediction * prediction, (1, 2))
a_01 = torch.sum(mask * prediction, (1, 2))
a_11 = torch.sum(mask, (1, 2))
# right hand side: b = [b_0, b_1]
b_0 = torch.sum(mask * prediction * target, (1, 2))
b_1 = torch.sum(mask * target, (1, 2))
# solution: x = A^-1 . b = [[a_11, -a_01], [-a_10, a_00]] / (a_00 * a_11 - a_01 * a_10) . b
x_0 = torch.zeros_like(b_0)
x_1 = torch.zeros_like(b_1)
det = a_00 * a_11 - a_01 * a_01
valid = det.nonzero()
x_0[valid] = (a_11[valid] * b_0[valid] - a_01[valid] * b_1[valid]) / det[valid]
x_1[valid] = (-a_01[valid] * b_0[valid] + a_00[valid] * b_1[valid]) / det[valid]
return x_0, x_1
def reduction_image_based(image_loss, M):
# mean of average of valid pixels of an image
# avoid division by 0 (if M = sum(mask) = 0: image_loss = 0)
valid = M.nonzero()
image_loss[valid] = image_loss[valid] / M[valid]
return torch.mean(image_loss)
def gradient_loss(prediction, target, mask, reduction=reduction_image_based):
M = torch.sum(mask, (1, 2))
diff = prediction - target
diff = torch.mul(mask, diff)
grad_x = torch.abs(diff[:, :, 1:] - diff[:, :, :-1])
mask_x = torch.mul(mask[:, :, 1:], mask[:, :, :-1])
grad_x = torch.mul(mask_x, grad_x)
grad_y = torch.abs(diff[:, 1:, :] - diff[:, :-1, :])
mask_y = torch.mul(mask[:, 1:, :], mask[:, :-1, :])
grad_y = torch.mul(mask_y, grad_y)
image_loss = torch.sum(grad_x, (1, 2)) + torch.sum(grad_y, (1, 2))
return reduction(image_loss, M)
def depth2pcd_fromplane(depth2d, c2w, K, h, w):
# pixel location
i, j = torch.meshgrid(torch.linspace(0, w - 1, w), torch.linspace(0, h - 1, h),
indexing='ij') # pytorch's meshgrid has indexing='ij'
i = i.t().cuda() # yy
j = j.t().cuda() # xx
x_normalized = (i - K[0][2]) / K[0][0]
y_normalized = (j - K[1][2]) / K[1][1]
# camera space
x_camera = x_normalized * depth2d
y_camera = y_normalized * depth2d
z_camera = depth2d
xyz1_camera = torch.concat([x_camera, y_camera, z_camera, torch.ones_like(x_camera)], 0).reshape(4,h*w) #
# print(c2w.shape, xyz1_camera.shape)
xyz_fine = torch.matmul(c2w, xyz1_camera).reshape(1,3,h,w) # (1 3 h w)
return xyz_fine
def least_square_normal_regress_fast01(x_depth3d, global_eye, global_b, size=9, gamma=0.15, depth_scaling_factor=1, eps=1e-5):
stride=1
xyz_padded = F.pad(x_depth3d, (size//2, size//2, size//2, size//2), mode='replicate')
xyz_patches = xyz_padded.unfold(2, size, stride).unfold(3, size, stride) # [batch_size, 3, width, height, size, size]
xyz_patches = xyz_patches.reshape((*xyz_patches.shape[:-2], ) + (-1,)) # [batch_size, 3, width, height, size*size]
xyz_perm = xyz_patches.permute([0, 2, 3, 4, 1])
diffs = xyz_perm - xyz_perm[:, :, :, (size*size)//2].unsqueeze(3)
diffs = diffs / xyz_perm[:, :, :, (size*size)//2].unsqueeze(3)
diffs[..., 0] = diffs[..., 2]
diffs[..., 1] = diffs[..., 2]
xyz_perm[torch.abs(diffs) > gamma] = 0.0
batch_size, width, height, size, _ = xyz_perm.shape
# Manual pseudoinverse
xyz_perm = xyz_perm.view(batch_size * width * height, size, 3)
A = torch.bmm(xyz_perm.transpose(1,2), xyz_perm)
A_det = torch.det(A)
A[A_det < eps, :, :] = global_eye
A_inv = torch.inverse(A) #.view(-1,3,3)
lstsq = A_inv.matmul(xyz_perm.transpose(1,2)).matmul(global_b).squeeze() # hw, 3
lstsq = torch.nn.functional.normalize(lstsq, dim=1)
lstsq[torch.isnan(lstsq)] = 0.0
return -lstsq.view([1,height,width,3]).permute([0, 3, 1, 2])
try:
from torch.utils.tensorboard import SummaryWriter
TENSORBOARD_FOUND = True
print("found tf board")
except ImportError:
TENSORBOARD_FOUND = False
print("not found tf board")
def saveRuntimeCode(dst: str) -> None:
additionalIgnorePatterns = ['.git', '.gitignore']
ignorePatterns = set()
ROOT = '.'
with open(os.path.join(ROOT, '.gitignore')) as gitIgnoreFile:
for line in gitIgnoreFile:
if not line.startswith('#'):
if line.endswith('\n'):
line = line[:-1]
if line.endswith('/'):
line = line[:-1]
ignorePatterns.add(line)
ignorePatterns = list(ignorePatterns)
for additionalPattern in additionalIgnorePatterns:
ignorePatterns.append(additionalPattern)
log_dir = pathlib.Path(__file__).parent.resolve()
shutil.copytree(log_dir, dst, ignore=shutil.ignore_patterns(*ignorePatterns))
print('Backup Finished!')
def training(dataset, opt, pipe, dataset_name, testing_iterations, saving_iterations, checkpoint_iterations, checkpoint, debug_from, wandb=None, logger=None, ply_path=None):
first_iter = 0
tb_writer = prepare_output_and_logger(dataset)
gaussians = GaussianModel(dataset.feat_dim, dataset.n_offsets, dataset.voxel_size, dataset.update_depth, dataset.update_init_factor, dataset.update_hierachy_factor, dataset.use_feat_bank, opt.attn_head_num, opt.attn_head_dim, use_bidirectional_attn=True)
scene = Scene(dataset, gaussians, ply_path=ply_path) # dataloader
gaussians.training_setup(opt)
if checkpoint:
(model_params, first_iter) = torch.load(checkpoint)
gaussians.restore(model_params, opt)
iter_start = torch.cuda.Event(enable_timing = True)
iter_end = torch.cuda.Event(enable_timing = True)
### load ref
all_image_names = [i for i in sorted(os.listdir(dataset.source_path+'/images_4')) if i.endswith('png')]
ref_name = all_image_names[40+29][:-4] # for spinnerf dataset
viewpoint_stack_first = scene.getTrainCameras().copy()
viewpoint_ref = [vp for vp in viewpoint_stack_first if vp.image_name == ref_name]
assert len(viewpoint_ref) == 1, len(viewpoint_ref)
viewpoint_ref = viewpoint_ref[0]
# load ref image
ref_image_path = dataset.ref_image_path
assert os.path.exists(ref_image_path), ref_image_path
assert ref_image_path.endswith('png'), ref_image_path
ref_image = torch.from_numpy(cv2.imread(ref_image_path)[:,:,[2,1,0]].astype(np.float32)) / 255.
ref_image = ref_image.permute(2, 0, 1).cuda() # read unaligned depth
# load ref depth
ref_depth_path = dataset.ref_depth_path
assert os.path.exists(ref_depth_path), ref_depth_path
assert ref_depth_path.endswith('npy'), ref_depth_path
ref_depth = torch.from_numpy(np.load(ref_depth_path).astype(np.float32)).unsqueeze(0).cuda() # read unaligned depth
# load K
camdata = read_cameras_binary(os.path.join(dataset.source_path, 'sparse/0/cameras.bin'))
H = camdata[1].height
W = camdata[1].width
img_wh = [W/4, H/4]
h, w = int(img_wh[1]), int(img_wh[0])
focal = camdata[1].params[0] * img_wh[0]/W
K = np.array([
[focal, 0, 0.5 * img_wh[0]],
[0, focal, 0.5 * img_wh[1]],
[0, 0, 1]
])
K = torch.from_numpy(K.astype(np.float32)).cuda()
print('K: ', K)
#### load ref
viewpoint_stack = None
ema_loss_for_log = 0.0
progress_bar = tqdm(range(first_iter, opt.iterations), desc="Training progress")
first_iter += 1
torch.autograd.set_detect_anomaly(True)
for iteration in range(first_iter, opt.iterations + 1):
# network gui not available in scaffold-gs yet
if network_gui.conn == None:
network_gui.try_connect()
while network_gui.conn != None:
try:
net_image_bytes = None
custom_cam, do_training, pipe.convert_SHs_python, pipe.compute_cov3D_python, keep_alive, scaling_modifer = network_gui.receive()
if custom_cam != None:
net_image = render(custom_cam, gaussians, pipe, background, scaling_modifer)["render"]
net_image_bytes = memoryview((torch.clamp(net_image, min=0, max=1.0) * 255).byte().permute(1, 2, 0).contiguous().cpu().numpy())
network_gui.send(net_image_bytes, dataset.source_path)
if do_training and ((iteration < int(opt.iterations)) or not keep_alive):
break
except Exception as e:
network_gui.conn = None
iter_start.record()
gaussians.update_learning_rate(iteration)
bg_color = [1, 1, 1] if dataset.white_background else [0, 0, 0]
background = torch.tensor(bg_color, dtype=torch.float32, device="cuda")
# Pick a random Camera
if not viewpoint_stack:
viewpoint_stack = scene.getTrainCameras().copy()
viewpoint_cam = viewpoint_stack.pop(randint(0, len(viewpoint_stack)-1))
# Render
if (iteration - 1) == debug_from:
pipe.debug = True
# gt preparation
gt_image = viewpoint_cam.original_image.cuda() # 3,h,w
gt_mask = viewpoint_cam.original_mask.cuda() # 1,h,w, fg = 1., bg = 0.
gt_image_name = viewpoint_cam.image_name
if gt_image_name == ref_name:
midas_depth = ref_depth
gt_image = ref_image
else:
midas_depth = viewpoint_cam.midas_depth.cuda()
# get mask:
voxel_visible_mask, position2D_x, position2D_y = prefilter_position2D(viewpoint_cam, gaussians, pipe, background)
# valid2D_mask
valid_mask_y = torch.logical_and(position2D_y>0, position2D_y<h)
valid_mask_x = torch.logical_and(position2D_x>0, position2D_x<w)
valid_mask_2D = torch.logical_and(valid_mask_x, valid_mask_y)
valid_mask_2D = torch.logical_and(valid_mask_2D, voxel_visible_mask) # anchor points, which splatting to this view
position2D_y = position2D_y.long()
position2D_x = position2D_x.long()
# if ((opt.enable_crossattn_refview > 0) and (gt_image_name == ref_name) and iteration > 15000) \
# or ((opt.enable_crossattn_otherview > 0) and (gt_image_name != ref_name) and iteration > 15000):
if ((opt.enable_crossattn_refview > 0) and (gt_image_name == ref_name) and iteration > opt.start_crossattn_from) \
or ((opt.enable_crossattn_otherview > 0) and (gt_image_name != ref_name) and iteration > opt.start_crossattn_from):
try:
# if True:
if opt.enable_edge_samping > 0:
min_y, max_y, min_x, max_x = sample_patch_in_mask_region(gt_mask=gt_mask[0], patch_size=256, small_ratio=opt.sampling_2D_small_ratio, max_h=h, max_w=w)
else:
print('No sampling method. ')
sampled_mask_2D = torch.zeros_like(gt_mask[0]).long()
sampled_mask_2D[min_y:max_y, min_x:max_x] = 1.0
anchor_2Dlabel_sample = sampled_mask_2D[position2D_y[valid_mask_2D], position2D_x[valid_mask_2D]] # (projected) anchor's 2D label
num_sampled = anchor_2Dlabel_sample.sum()
if not (num_sampled > 0):
print('num_sampled <= 0, %d'%num_sampled)
exit()
voxel_sampled_mask = -1 * torch.ones_like(voxel_visible_mask) # -1
voxel_sampled_mask[valid_mask_2D] = anchor_2Dlabel_sample # -1: invalid, 0: unampled, 1: sampled
anchor_2Dlabel_fgbg = gt_mask[0].long()[position2D_y[valid_mask_2D], position2D_x[valid_mask_2D]] # (projected) anchor's 2D label
voxel_fgbg_mask = -1 * torch.ones_like(voxel_visible_mask) # -1
voxel_fgbg_mask[valid_mask_2D] = anchor_2Dlabel_fgbg # -1: invalid, 0: bg, 1: fg
# 2-category cluster
fgbg_sampled = voxel_fgbg_mask[voxel_sampled_mask>0] # P
anchor_sampled = gaussians.get_anchor[voxel_sampled_mask>0] # P, 3
anchor_sampled_fg_xyz = anchor_sampled[fgbg_sampled>0] # P1 x 3
anchor_sampled_fg_indices = torch.nonzero(fgbg_sampled>0).squeeze() # P1
anchor_sampled_bg_xyz = anchor_sampled[fgbg_sampled==0] # P2 x 3
anchor_sampled_bg_indices = torch.nonzero(fgbg_sampled==0).squeeze() # P2
if (anchor_sampled.shape[0] <= 11):
print('anchor_sampled <= 11, %d'%num_sampled)
exit()
if (anchor_sampled_fg_xyz.shape[0] <= 11):
print('anchor_sampled <= 11, %d'%num_sampled)
exit()
if (anchor_sampled_bg_xyz.shape[0] <= 11):
print('anchor_sampled <= 11, %d'%num_sampled)
exit()
row_indices = anchor_sampled_fg_indices
sampled_indices = anchor_sampled_bg_indices
if not (sampled_indices.shape[0] > 0):
print('number pair <=0, %d'%sampled_indices.shape[0])
exit()
min_num = min(min(sampled_indices.shape[0], row_indices.shape[0]), 2000)
unique_sampled_indices = sampled_indices[torch.randperm(sampled_indices.size(0))][:min_num]
unique_row_indices = row_indices[torch.randperm(row_indices.size(0))][:min_num]
voxel_sampled_src = torch.zeros_like(voxel_visible_mask) # P
voxel_used_src = torch.zeros(num_sampled).cuda().bool()
voxel_used_src[unique_row_indices] = True
voxel_sampled_src[voxel_sampled_mask>0] = voxel_used_src
voxel_sampled_dst = torch.zeros_like(vo
10BC8
xel_visible_mask)
voxel_used_dst = torch.zeros(num_sampled).cuda().bool()
voxel_used_dst[unique_sampled_indices] = True
voxel_sampled_dst[voxel_sampled_mask>0] = voxel_used_dst # P
# 3. choose
gaussians.run_crossattn(voxel_sampled_src, voxel_sampled_dst, pe=(opt.enable_pe>0), ema=opt.crossattn_feat_update_ema, is_ref=(gt_image_name == ref_name))
cross_flag = True
except:
# else:
print('No valid sampled anchors...')
cross_flag = False
continue
else:
cross_flag = False
retain_grad = (iteration < opt.update_until and iteration >= 0)
# render with mask
render_pkg = render(viewpoint_cam, gaussians, pipe, background, visible_mask=voxel_visible_mask, retain_grad=retain_grad)
image, viewspace_point_tensor, visibility_filter, offset_selection_mask, radii, scaling, opacity =\
render_pkg["render"], render_pkg["viewspace_points"], render_pkg["visibility_filter"], render_pkg["selection_mask"], render_pkg["radii"], render_pkg["scaling"], render_pkg["neural_opacity"]
# render_uncertainty = render_pkg["uncertainty"]
# RGB loss
if gt_image_name == ref_name:
print('Using reference view... ')
Ll1 = l1_loss(image, gt_image)
loss = opt.refer_rgb_lr * ((1.0 - opt.lambda_dssim) * Ll1 + opt.lambda_dssim * (1.0 - ssim(image, gt_image)))
if opt.refer_rgb_lr_fg > opt.refer_rgb_lr:
loss += (opt.refer_rgb_lr_fg - opt.refer_rgb_lr) * ((1.0 - opt.lambda_dssim) * l1_loss_masked(image, gt_image, gt_mask) + opt.lambda_dssim * (1.0 - ssim_masked(image, gt_image, gt_mask)))
else:
other_rgb_weight = (1.0-gt_mask) + opt.other_rgb_lr_fg * gt_mask
Ll1 = l1_loss_masked(image, gt_image, other_rgb_weight)
loss = opt.other_rgb_lr * ((1.0 - opt.lambda_dssim) * Ll1 + opt.lambda_dssim * (1.0 - ssim_masked(image, gt_image, other_rgb_weight)))
# Depth loss
if opt.refer_depth_lr > 0 and gt_image_name == ref_name:
depth = render_pkg["render_depth"]
valid_mask = 1.0-gt_mask
scale, shift = compute_scale_and_shift(depth, midas_depth, valid_mask)
scale = torch.abs(scale) # for excluding negative scale
aligned_depth = scale.view(-1, 1, 1) * depth + shift.view(-1, 1, 1)
loss += opt.refer_depth_lr * l1_loss(aligned_depth, midas_depth)
if opt.refer_depth_lr_fg > opt.refer_depth_lr:
fg_mask = get_random_mask(gt_mask[0], enlarge_ratio=1.5, small_ratio=0.8, max_h=h, max_w=w)
loss += (opt.refer_depth_lr_fg - opt.refer_depth_lr) * l1_loss_masked(aligned_depth, midas_depth, fg_mask)
for scale in range(4):
step = pow(2, scale)
loss += 0.5 * opt.refer_depth_lr_smooth * gradient_loss(aligned_depth[:,::step,::step], midas_depth[:,::step,::step], torch.ones_like(gt_mask)[:,::step,::step])
# 2.2 depth loss from other view
if opt.other_depth_lr > 0 and gt_image_name != ref_name:
depth = render_pkg["render_depth"]
valid_mask = (1.0-gt_mask)
scale, shift = compute_scale_and_shift(depth, midas_depth, valid_mask)
scale = torch.abs(scale) # modi here
aligned_depth = scale.view(-1, 1, 1) * depth + shift.view(-1, 1, 1)
loss += opt.other_depth_lr * l1_loss_masked(aligned_depth, midas_depth, valid_mask)
for scale in range(4):
step = pow(2, scale)
# loss += 0.5 * opt.other_depth_lr * gradient_loss(aligned_depth[:,::step,::step], midas_depth[:,::step,::step], (valid_mask)[:,::step,::step])
loss += 0.5 * opt.other_depth_lr_smooth * gradient_loss(aligned_depth[:,::step,::step], midas_depth[:,::step,::step], (valid_mask)[:,::step,::step])
loss.backward(retain_graph=True) # 200
iter_end.record()
with torch.no_grad():
# Progress bar
ema_loss_for_log = 0.4 * loss.item() + 0.6 * ema_loss_for_log
if iteration % 10 == 0:
progress_bar.set_postfix({"Loss": f"{ema_loss_for_log:.{7}f}"})
progress_bar.update(10)
if iteration == opt.iterations:
progress_bar.close()
# Log and save
training_report(tb_writer, dataset_name, iteration, Ll1, loss, l1_loss, iter_start.elapsed_time(iter_end), testing_iterations, scene, render, (pipe, background), wandb, logger)
if (iteration in saving_iterations):
logger.info("\n[ITER {}] Saving Gaussians".format(iteration))
scene.save(iteration)
# densification
if iteration < opt.update_until and iteration > opt.start_stat: # 500 - 15000
# add statis
gaussians.training_statis(viewspace_point_tensor, opacity, visibility_filter, offset_selection_mask, voxel_visible_mask)
# densification
if iteration > opt.update_from and iteration % opt.update_interval == 0: # > 1500， 100
gaussians.adjust_anchor(check_interval=opt.update_interval, success_threshold=opt.success_threshold, grad_threshold=opt.densify_grad_threshold, min_opacity=opt.min_opacity)
elif iteration == opt.update_until:
del gaussians.opacity_accum
del gaussians.offset_gradient_accum
del gaussians.offset_denom
torch.cuda.empty_cache()
# Optimizer step
if iteration < opt.iterations:
gaussians.optimizer.step()
gaussians.optimizer.zero_grad(set_to_none = True)
if cross_flag:
gaussians.optimizer_c.step()
gaussians.optimizer_c.zero_grad(set_to_none = True)
if (iteration in checkpoint_iterations):
logger.info("\n[ITER {}] Saving Checkpoint".format(iteration))
torch.save((gaussians.capture(), iteration), scene.model_path + "/chkpnt" + str(iteration) + ".pth")
def prepare_output_and_logger(args):
if not args.model_path:
if os.getenv('OAR_JOB_ID'):
unique_str=os.getenv('OAR_JOB_ID')
else:
unique_str = str(uuid.uuid4())
args.model_path = os.path.join("./output/", unique_str[0:10])
# Set up output folder
print("Output folder: {}".format(args.model_path))
os.makedirs(args.model_path, exist_ok = True)
with open(os.path.join(args.model_path, "cfg_args"), 'w') as cfg_log_f:
cfg_log_f.write(str(Namespace(**vars(args))))
# Create Tensorboard writer
tb_writer = None
if TENSORBOARD_FOUND:
tb_writer = SummaryWriter(args.model_path)
else:
print("Tensorboard not available: not logging progress")
return tb_writer
def training_report(tb_writer, dataset_name, iteration, Ll1, loss, l1_loss, elapsed, testing_iterations, scene : Scene, renderFunc, renderArgs, wandb=None, logger=None):
if tb_writer:
tb_writer.add_scalar(f'{dataset_name}/train_loss_patches/l1_loss', Ll1.item(), iteration)
tb_writer.add_scalar(f'{dataset_name}/train_loss_patches/total_loss', loss.item(), iteration)
tb_writer.add_scalar(f'{dataset_name}/iter_time', elapsed, iteration)
if wandb is not None:
wandb.log({"train_l1_loss":Ll1, 'train_total_loss':loss, })
# Report test and samples of training set
if iteration in testing_iterations:
scene.gaussians.eval()
torch.cuda.empty_cache()
validation_configs = ({'name': 'test', 'cameras' : scene.getTestCameras()},
{'name': 'train', 'cameras' : [scene.getTrainCameras()[idx % len(scene.getTrainCameras())] for idx in range(5, 30, 5)]})
for config in validation_configs:
if config['cameras'] and len(config['cameras']) > 0:
l1_test = 0.0
psnr_test = 0.0
if wandb is not None:
gt_image_list = []
render_image_list = []
errormap_list = []
for idx, viewpoint in enumerate(config['cameras']):
voxel_visible_mask, _, _ = prefilter_position2D(viewpoint, scene.gaussians, *renderArgs)
image = torch.clamp(renderFunc(viewpoint, scene.gaussians, *renderArgs, visible_mask=voxel_visible_mask)["render"], 0.0, 1.0)
gt_image = torch.clamp(viewpoint.original_image.to("cuda"), 0.0, 1.0)
if tb_writer and (idx < 30):
tb_writer.add_images(f'{dataset_name}/'+config['name'] + "_view_{}/render".format(viewpoint.image_name), image[None], global_step=iteration)
tb_writer.add_images(f'{dataset_name}/'+config['name'] + "_view_{}/errormap".format(viewpoint.image_name), (gt_image[None]-image[None]).abs(), global_step=iteration)
if wandb:
render_image_list.append(image[None])
errormap_list.append((gt_image[None]-image[None]).abs())
if iteration == testing_iterations[0]:
tb_writer.add_images(f'{dataset_name}/'+config['name'] + "_view_{}/ground_truth".format(viewpoint.image_name), gt_image[None], global_step=iteration)
if wandb:
gt_image_list.append(gt_image[None])
l1_test += l1_loss(image, gt_image).mean().double()
psnr_test += psnr(image, gt_image).mean().double()
psnr_test /= len(config['cameras'])
l1_test /= len(config['cameras'])
logger.info("\n[ITER {}] Evaluating {}: L1 {} PSNR {}".format(iteration, config['name'], l1_test, psnr_test))
if tb_writer:
tb_writer.add_scalar(f'{dataset_name}/'+config['name'] + '/loss_viewpoint - l1_loss', l1_test, iteration)
tb_writer.add_scalar(f'{dataset_name}/'+config['name'] + '/loss_viewpoint - psnr', psnr_test, iteration)
if wandb is not None:
wandb.log({f"{config['name']}_loss_viewpoint_l1_loss":l1_test, f"{config['name']}_PSNR":psnr_test})
if tb_writer:
tb_writer.add_scalar(f'{dataset_name}/'+'total_points', scene.gaussians.get_anchor.shape[0], iteration)
torch.cuda.empty_cache()
scene.gaussians.train()
def render_set(model_path, name, iteration, views, gaussians, pipeline, background, source_path=None):
render_path = os.path.join(model_path, name, "ours_{}".format(iteration), "renders")
if name !='spiral':
error_path = os.path.join(model_path, name, "ours_{}".format(iteration), "errors")
gts_path = os.path.join(model_path, name, "ours_{}".format(iteration), "gt")
else:
rgbd_path = os.path.join(model_path, name, "ours_{}".format(iteration), "rgbd")
depth_path = os.path.join(model_path, name, "ours_{}".format(iteration), "depth_renders")
uncertainty_path = os.path.join(model_path, name, "ours_{}".format(iteration), "uncertainty_renders")
makedirs(render_path, exist_ok=True)
if name !='spiral':
makedirs(error_path, exist_ok=True)
makedirs(gts_path, exist_ok=True)
else:
makedirs(rgbd_path, exist_ok=True)
makedirs(depth_path, exist_ok=True)
makedirs(uncertainty_path, exist_ok=True)
if name == 'spiral':
RGBD = []
camdata = read_cameras_binary(os.path.join(source_path, 'sparse/0/cameras.bin'))
H = camdata[1].height
W = camdata[1].width
img_wh = [W/4, H/4]
h, w = int(img_wh[1]), int(img_wh[0])
focal = camdata[1].params[0] * img_wh[0]/W
K = np.array([
[focal, 0, 0.5 * img_wh[0]],
[0, focal, 0.5 * img_wh[1]],
[0, 0, 1]
])
K = torch.from_numpy(K.astype(np.float32)).cuda()
print('K: ', K)
# for norm calculation
kernel_size = 9
global_eye = torch.eye(3).cuda()
global_b = torch.ones([h*w,kernel_size**2,1]).cuda()
t_list = []
visible_count_list = []
name_list = []
per_view_dict = {}
for idx, view in enumerate(tqdm(views, desc="Rendering progress")):
torch.cuda.synchronize();t_start = time.time()
# voxel_visible_mask = prefilter_voxel(view, gaussians, pipeline, background)
voxel_visible_mask, _, _ = prefilter_position2D(view, gaussians, pipeline, background)
render_pkg = render(view, gaussians, pipeline, background, visible_mask=voxel_visible_mask)
torch.cuda.synchronize();t_end = time.time()
t_list.append(t_end - t_start)
# renders
rendering = torch.clamp(render_pkg["render"], 0.0, 1.0)
visible_count = (render_pkg["radii"] > 0).sum()
visible_count_list.append(visible_count)
# uncertainty
uncertainty = torch.clamp(render_pkg["uncertainty"], 0.0, 1.0)
# add depth rendering
if name != 'spiral':
render_depth = render_pkg["render_depth"]
depth_ori = render_depth.clone()
# gt_depth = view.depth
gt_depth = view.midas_depth
gt_mask = view.original_mask
valid_mask = 1 - gt_mask
scale, shift = compute_scale_and_shift(render_depth, gt_depth, valid_mask)
scale = torch.abs(scale)
depth = render_depth * scale + shift
depth_concat = torch.cat((depth, gt_depth), dim=0).unsqueeze(1)
tensor = torchvision.utils.make_grid(depth_concat, padding=0, normalize=False, scale_each=False).cpu().detach().numpy()
plt.imsave(os.path.join(depth_path, '{0:05d}'.format(idx) + "_depth.png"), np.transpose(tensor, (1,2,0))[:,:,0], cmap="viridis")
# gts
gt = view.original_image[0:3, :, :]
# error maps
errormap = (rendering - gt).abs()
name_list.append('{0:05d}'.format(idx) + ".png")
torchvision.utils.save_image(rendering, os.path.join(render_path, '{0:05d}'.format(idx) + ".png"))
torchvision.utils.save_image(uncertainty, os.path.join(uncertainty_path, '{0:05d}'.format(idx) + ".png"))
torchvision.utils.save_image(errormap, os.path.join(error_path, '{0:05d}'.format(idx) + ".png"))
torchvision.utils.save_image(gt, os.path.join(gts_path, '{0:05d}'.format(idx) + ".png"))
per_view_dict['{0:05d}'.format(idx) + ".png"] = visible_count.item()
else:
depth = render_pkg["render_depth"]
depth_ori = depth.clone()
depth = (depth - depth.min()) / (depth.max() - depth.min())
depth = (255 * depth.cpu().numpy()).astype(np.uint8)[0]
depth = cv2.applyColorMap(depth, cv2.COLORMAP_INFERNO)[:,:,[2,1,0]]
depth = torch.FloatTensor(depth).permute([2,0,1]) / 255.
rendered = torch.flip(rendering.unsqueeze(0), dims=[3])[0]
uncertainty = torch.flip(uncertainty.unsqueeze(0), dims=[3])[0] # 1 h w
uncertainty = uncertainty.expand([3, -1, -1])
depth = torch.flip(depth.unsqueeze(0), dims=[3])[0]
# add normal
c2w = view.C2W[:3,:4]
c2w = torch.from_numpy(c2w.astype(np.float32)).cuda()
depth_map_3d = depth2pcd_fromplane(depth2d=depth_ori, c2w=c2w, K=K, h=h, w=w)
normal = least_square_normal_regress_fast01(depth_map_3d, global_eye, global_b)
normal = normal.reshape(3,h,w).permute([1,2,0])
normal = (torch.nn.functional.normalize(normal, p=2, dim=-1)).permute([2,0,1]) # 3, h, w
normal = (normal + 1) / 2
normal = torch.flip(normal.unsqueeze(0), dims=[3])[0]
rgbd = torch.concat([rendered.cpu(), depth, normal.cpu(), uncertainty.cpu()], 2)
torchvision.utils.save_image(rendered, os.path.join(render_path, '{0:05d}'.format(idx) + ".png"))
torchvision.utils.save_image(uncertainty, os.path.join(uncertainty_path, '{0:05d}'.format(idx) + ".png"))
if name != 'spiral':
torchvision.utils.save_image(gt, os.path.join(gts_path, '{0:05d}'.format(idx) + ".png"))
else:
torchvision.utils.save_image(rgbd, os.path.join(rgbd_path, '{0:05d}'.format(idx) + ".png"))
torchvision.utils.save_image(depth, os.path.join(depth_path, '{0:05d}'.format(idx) + ".png"))
if name == 'train':
np.save(os.path.join(depth_path, '{0:05d}'.format(idx) + ".npy"), depth_ori.cpu().numpy())
if name != 'spiral':
with open(os.path.join(model_path, name, "ours_{}".format(iteration), "per_view_count.json"), 'w') as fp:
json.dump(per_view_dict, fp, indent=True)
else:
cmd = f"/usr/local/bin/ffmpeg -y -i {rgbd_path}/%05d.png -q:v 0 {os.path.join(model_path, name,'-'.join(model_path.split('/')[1:-1])+'.mp4')}"
print(cmd)
os.system(cmd)
return t_list, visible_count_list
def render_sets(dataset : ModelParams, iteration : int, pipeline : PipelineParams, skip_train=False, skip_test=False, wandb=None, tb_writer=None, dataset_name=None, logger=None):
with torch.no_grad():
gaussians = GaussianModel(dataset.feat_dim, dataset.n_offsets, dataset.voxel_size, dataset.update_depth, dataset.update_init_factor, dataset.update_hierachy_factor, dataset.use_feat_bank)
dataset.pretrained_model_path = ""
scene = Scene(dataset, gaussians, load_iteration=iteration, shuffle=False)
gaussians.eval()
bg_color = [1,1,1] if dataset.white_background else [0, 0, 0]
background = torch.tensor(bg_color, dtype=torch.float32, device="cuda")
t_spiral_list, visible_count = render_set(dataset.model_path, "spiral", scene.loaded_iter, scene.getSpiralCameras(), gaussians, pipeline, background, dataset.source_path)
spiral_fps = 1.0 / torch.tensor(t_spiral_list[5:]).mean()
logger.info(f'Spiral FPS: \033[1;35m{spiral_fps.item():.5f}\033[0m')
if tb_writer:
tb_writer.add_scalar(f'{dataset_name}/spiral_FPS', spiral_fps.item(), 0)
if wandb is not None:
wandb.log({"spiral_fps":spiral_fps, })
if not skip_train:
t_train_list, _ = render_set(dataset.model_path, "train", scene.loaded_iter, scene.getTrainCameras(), gaussians, pipeline, background)
train_fps = 1.0 / torch.tensor(t_train_list[5:]).mean()
logger.info(f'Train FPS: \033[1;35m{train_fps.item():.5f}\033[0m')
if wandb is not None:
wandb.log({"train_fps":train_fps.item(), })
if not skip_test:
t_test_list, visible_count = render_set(dataset.model_path, "test", scene.loaded_iter, scene.getTestCameras(), gaussians, pipeline, background)
test_fps = 1.0 / torch.tensor(t_test_list[5:]).mean()
logger.info(f'Test FPS: \033[1;35m{test_fps.item():.5f}\033[0m')
if tb_writer:
tb_writer.add_scalar(f'{dataset_name}/test_FPS', test_fps.item(), 0)
if wandb is not None:
wandb.log({"test_fps":test_fps, })
return visible_count
def readImages(renders_dir, gt_dir, mask_dir):
renders = []
gts = []
image_names = []
masks = []
for fname in os.listdir(renders_dir):
render = Image.open(renders_dir / fname)
gt = Image.open(gt_dir / fname)
renders.append(tf.to_tensor(render).unsqueeze(0)[:, :3, :, :].cuda())
gts.append(tf.to_tensor(gt).unsqueeze(0)[:, :3, :, :].cuda())
image_names.append(fname)
mask = Image.open(os.path.join(mask_dir, 'out_%05d.png'%(1+int(fname.split('.')[0]))))
mask = mask.resize((1008, 567), Image.LANCZOS)
masks.append(tf.to_tensor(mask).bool().expand([3,-1,-1]).unsqueeze(0).cuda())
return renders, gts, image_names, masks
def evaluate(dataset, model_paths, visible_count=None, wandb=None, tb_writer=None, dataset_name=None, logger=None):
full_dict = {}
per_view_dict = {}
full_dict_polytopeonly = {}
per_view_dict_polytopeonly = {}
print("")
scene_dir = model_paths
full_dict[scene_dir] = {}
per_view_dict[scene_dir] = {}
full_dict_polytopeonly[scene_dir] = {}
per_view_dict_polytopeonly[scene_dir] = {}
test_dir = Path(scene_dir) / "test"
mask_dir = Path(dataset.source_path) / "images_4" / "test_label"
for method in os.listdir(test_dir):
full_dict[scene_dir][method] = {}
per_view_dict[scene_dir][method] = {}
full_dict_polytopeonly[scene_dir][method] = {}
per_view_dict_polytopeonly[scene_dir][method] = {}
method_dir = test_dir / method
gt_dir = method_dir/ "gt"
renders_dir = method_dir / "renders"
renders, gts, image_names, masks = readImages(renders_dir, gt_dir, mask_dir)
ssims = []
psnrs = []
lpipss = []
masked_ssims = []
masked_psnrs = []
masked_lpipss = []
for idx in tqdm(range(len(renders)), desc="Metric evaluation progress"):
# print(renders[idx].shape, renders[idx].dtype, renders[idx].min(), renders[idx].max())
# print(gts[idx].shape, gts[idx].dtype, gts[idx].min(), gts[idx].max())
ssims.append(my_ssim(renders[idx], gts[idx]))
psnrs.append(psnr(renders[idx], gts[idx]))
lpipss.append(lpips_fn(2*renders[idx]-1, 2*gts[idx]-1).detach())
masked_ssims.append(my_ssim(renders[idx], gts[idx], masks[idx]))
masked_psnrs.append(psnr(renders[idx], gts[idx], masks[idx]))
cur_lpips_spatial = loss_fn_vgg_spatial(2*renders[idx]-1, 2*gts[idx]-1).squeeze()
masked_lpipss.append(cur_lpips_spatial[masks[idx][0,0]].mean())
if wandb is not None:
wandb.log({"test_SSIMS":torch.stack(ssims).mean().item(), })
wandb.log({"test_PSNR_final":torch.stack(psnrs).mean().item(), })
wandb.log({"test_LPIPS":torch.stack(lpipss).mean().item(), })
wandb.log({"test_masked_SSIMS":torch.stack(masked_ssims).mean().item(), })
wandb.log({"test_masked_PSNR_final":torch.stack(masked_psnrs).mean().item(), })
wandb.log({"test_masked_LPIPS":torch.stack(masked_lpipss).mean().item(), })
logger.info(f"model_paths: \033[1;35m{model_paths}\033[0m")
logger.info(" SSIM : \033[1;35m{:>12.7f}\033[0m".format(torch.tensor(ssims).mean(), ".5"))
logger.info(" PSNR : \033[1;35m{:>12.7f}\033[0m".format(torch.tensor(psnrs).mean(), ".5"))
logger.info(" LPIPS: \033[1;35m{:>12.7f}\033[0m".format(torch.tensor(lpipss).mean(), ".5"))
logger.info(" masked SSIM : \033[1;35m{:>12.7f}\033[0m".format(torch.tensor(masked_ssims).mean(), ".5"))
logger.info(" masked PSNR : \033[1;35m{:>12.7f}\033[0m".format(torch.tensor(masked_psnrs).mean(), ".5"))
logger.info(" masked LPIPS: \033[1;35m{:>12.7f}\033[0m".format(torch.tensor(masked_lpipss).mean(), ".5"))
print("")
if tb_writer:
tb_writer.add_scalar(f'{dataset_name}/SSIM', torch.tensor(ssims).mean().item(), 0)
tb_writer.add_scalar(f'{dataset_name}/PSNR', torch.tensor(psnrs).mean().item(), 0)
tb_writer.add_scalar(f'{dataset_name}/LPIPS', torch.tensor(lpipss).mean().item(), 0)
tb_writer.add_scalar(f'{dataset_name}/VISIBLE_NUMS', torch.tensor(visible_count).mean().item(), 0)
full_dict[scene_dir][method].update({"SSIM": torch.tensor(ssims).mean().item(),
"PSNR": torch.tensor(psnrs).mean().item(),
"LPIPS": torch.tensor(lpipss).mean().item()})
per_view_dict[scene_dir][method].update({"SSIM": {name: ssim for ssim, name in zip(torch.tensor(ssims).tolist(), image_names)},
"PSNR": {name: psnr for psnr, name in zip(torch.tensor(psnrs).tolist(), image_names)},
"LPIPS": {name: lp for lp, name in zip(torch.tensor(lpipss).tolist(), image_names)},
"VISIBLE_COUNT": {name: vc for vc, name in zip(torch.tensor(visible_count).tolist(), image_names)}})
with open(scene_dir + "/results.json", 'w') as fp:
json.dump(full_dict[scene_dir], fp, indent=True)
with open(scene_dir + "/per_view.json", 'w') as fp:
json.dump(per_view_dict[scene_dir], fp, indent=True)
def get_logger(path):
import logging
logger = logging.getLogger()
logger.setLevel(logging.INFO)
fileinfo = logging.FileHandler(os.path.join(path, "outputs.log"))
fileinfo.setLevel(logging.INFO)