-
Notifications
You must be signed in to change notification settings - Fork 8
/
test.py
374 lines (311 loc) · 14.5 KB
/
test.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
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
# Copyright (c) Gorilla-Lab. All rights reserved.
import argparse
import numpy as np
import os
import torch
import spconv
import scipy.stats as stats
import pointgroup_ops
import gorilla
import gorilla3d
import gorilla3d.datasets as datasets
import sstnet
def get_parser():
parser = argparse.ArgumentParser(description="SSTNet for Point Cloud Instance Segmentation")
parser.add_argument("--config",
type=str,
default="config/default.yaml",
help="path to config file")
### pretrain
parser.add_argument("--pretrain",
type=str,
default="",
help="path to pretrain model")
### split
parser.add_argument("--split",
type=str,
default="val",
help="dataset split to test")
### semantic only
parser.add_argument("--semantic",
action="store_true",
help="only evaluate semantic segmentation")
### log file path
parser.add_argument("--log-file",
type=str,
default=None,
help="log_file path")
### test srcipt operation
parser.add_argument("--eval",
action="store_true",
help="evaluate or not")
parser.add_argument("--save",
action="store_true",
help="save results or not")
parser.add_argument("--visual",
type=str,
default=None,
help="visual path, give to save visualization results")
args_cfg = parser.parse_args()
return args_cfg
def init():
args = get_parser()
cfg = gorilla.Config.fromfile(args.config)
cfg.pretrain = args.pretrain
cfg.semantic = args.semantic
cfg.dataset.task = args.split # change tasks
cfg.data.visual = args.visual
cfg.data.eval = args.eval
cfg.data.save = args.save
gorilla.set_random_seed(cfg.data.test_seed)
#### get logger file
params_dict = dict(
epoch=cfg.data.test_epoch,
optim=cfg.optimizer.type,
lr=cfg.optimizer.lr,
scheduler=cfg.lr_scheduler.type
)
if "test" in args.split:
params_dict["suffix"] = "test"
log_dir, logger = gorilla.collect_logger(
prefix=os.path.splitext(args.config.split("/")[-1])[0],
log_name="test",
log_file=args.log_file,
# **params_dict
)
logger.info(
"************************ Start Logging ************************")
# log the config
logger.info(cfg)
global result_dir
result_dir = os.path.join(
log_dir, "result",
"epoch{}_nmst{}_scoret{}_npointt{}".format(cfg.data.test_epoch,
cfg.data.TEST_NMS_THRESH,
cfg.data.TEST_SCORE_THRESH,
cfg.data.TEST_NPOINT_THRESH),
args.split)
os.makedirs(os.path.join(result_dir, "predicted_masks"), exist_ok=True)
global semantic_label_idx
os.environ["CUDA_VISIBLE_DEVICES"] = "1"
semantic_label_idx = torch.tensor([
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 16, 24, 28, 33, 34, 36, 39
]).cuda()
return logger, cfg
def test(model, cfg, logger):
logger.info(">>>>>>>>>>>>>>>> Start Evaluation >>>>>>>>>>>>>>>>")
epoch = cfg.data.test_epoch
semantic = cfg.semantic
cfg.dataset.test_mode = True
cfg.dataloader.batch_size = 1
cfg.dataloader.num_workers = 2
test_dataset = gorilla.build_dataset(cfg.dataset)
test_dataloader = gorilla.build_dataloader(test_dataset, cfg.dataloader)
with torch.no_grad():
model = model.eval()
# init timer to calculate time
timer = gorilla.Timer()
# define evaluator
# get the real data root
data_root = os.path.join(os.path.dirname(__file__), cfg.dataset.data_root)
sub_dir = "scans_test" if "test" in cfg.dataset.task else "scans"
label_root = os.path.join(data_root, cfg.dataset.task + "_gt")
evaluator = gorilla3d.ScanNetSemanticEvaluator(label_root)
inst_evaluator = gorilla3d.ScanNetInstanceEvaluator(label_root)
for i, batch in enumerate(test_dataloader):
torch.cuda.empty_cache()
timer.reset()
N = batch["feats"].shape[0]
test_scene_name = batch["scene_list"][0]
coords = batch["locs"].cuda() # [N, 1 + 3] dimension 0 for batch_idx
locs_offset = batch["locs_offset"].cuda() # [B, 3]
voxel_coords = batch["voxel_locs"].cuda() # [M, 1 + 3]
p2v_map = batch["p2v_map"].cuda() # [N]
v2p_map = batch["v2p_map"].cuda() # [M, 1 + maxActive]
coords_float = batch["locs_float"].cuda() # [N, 3]
feats = batch["feats"].cuda() # [N, C]
batch_offsets = batch["offsets"].cuda() # [B + 1]
scene_list = batch["scene_list"]
superpoint = batch["superpoint"].cuda() # [N
_, superpoint = torch.unique(superpoint, return_inverse=True) # [N]
extra_data = {"batch_idxs": coords[:, 0].int(),
"superpoint": superpoint,
"locs_offset": locs_offset,
"scene_list": scene_list}
spatial_shape = batch["spatial_shape"]
if cfg.model.use_coords:
feats = torch.cat((feats, coords_float), 1)
voxel_feats = pointgroup_ops.voxelization(feats, v2p_map, cfg.data.mode) # [M, C]
input_ = spconv.SparseConvTensor(voxel_feats, voxel_coords.int(), spatial_shape, cfg.dataloader.batch_size)
data_time = timer.since_last()
ret = model(input_,
p2v_map,
coords_float,
epoch,
extra_data,
mode="test",
semantic_only=semantic)
semantic_scores = ret["semantic_scores"] # [N, nClass]
pt_offsets = ret["pt_offsets"] # [N, 3]
score_epochs = cfg.model.score_epochs
prepare_flag = epoch > score_epochs
if prepare_flag and not semantic:
scores = ret["proposal_scores"]
##### preds
with torch.no_grad():
preds = {}
preds["semantic"] = semantic_scores
preds["pt_offsets"] = pt_offsets
if prepare_flag and not semantic:
proposals_idx, proposals_offset = ret["proposals"]
preds["score"] = scores
preds["proposals"] = (proposals_idx, proposals_offset)
##### get predictions (#1 semantic_pred, pt_offsets; #2 scores, proposals_pred)
semantic_scores = preds["semantic"] # [N, nClass=20]
semantic_pred = semantic_scores.max(1)[1] # [N]
pt_offsets = preds["pt_offsets"] # [N, 3]
##### semantic segmentation evaluation
if cfg.data.eval:
inputs = [{"scene_name": test_scene_name}]
outputs = [{"semantic_pred": semantic_pred}]
evaluator.process(inputs, outputs)
if prepare_flag and not semantic:
scores = preds["score"] # [num_prop, 1]
scores_pred = torch.sigmoid(scores.view(-1))
proposals_idx, proposals_offset = preds["proposals"]
# proposals_idx: (sumNPoint, 2) dim 0 for cluster_id, dim 1 for corresponding point idxs in N
# proposals_offset: (num_prop + 1)
proposals_pred = torch.zeros(
(proposals_offset.shape[0] - 1, N),
dtype=torch.int,
device=scores_pred.device) # [num_prop, N]
proposals_pred[proposals_idx[:, 0].long(),
proposals_idx[:, 1].long()] = 1
semantic_pred_list = []
for start, end in zip(proposals_offset[:-1],
proposals_offset[1:]):
semantic_label, _ = stats.mode(
semantic_pred[proposals_idx[start:end,
1].long()].cpu().numpy())
semantic_label = semantic_label[0]
semantic_pred_list.append(semantic_label)
semantic_id = semantic_label_idx[semantic_pred_list]
##### score threshold
score_mask = (scores_pred > cfg.data.TEST_SCORE_THRESH)
scores_pred = scores_pred[score_mask]
proposals_pred = proposals_pred[score_mask]
semantic_id = semantic_id[score_mask]
##### npoint threshold
proposals_pointnum = proposals_pred.sum(1)
npoint_mask = (proposals_pointnum > cfg.data.TEST_NPOINT_THRESH)
scores_pred = scores_pred[npoint_mask]
proposals_pred = proposals_pred[npoint_mask]
semantic_id = semantic_id[npoint_mask]
##### nms
if semantic_id.shape[0] == 0:
pick_idxs = np.empty(0)
else:
proposals_pred_f = proposals_pred.float(
) # [num_prop, N], float, cuda
intersection = torch.mm(
proposals_pred_f, proposals_pred_f.t(
)) # [num_prop, num_prop], float, cuda
proposals_pointnum = proposals_pred_f.sum(
1) # [num_prop], float, cuda
proposals_pn_h = proposals_pointnum.unsqueeze(-1).repeat(
1, proposals_pointnum.shape[0])
proposals_pn_v = proposals_pointnum.unsqueeze(0).repeat(
proposals_pointnum.shape[0], 1)
cross_ious = intersection / (proposals_pn_h +
proposals_pn_v - intersection)
pick_idxs = gorilla3d.non_max_suppression(
cross_ious.cpu().numpy(),
scores_pred.cpu().numpy(),
cfg.data.TEST_NMS_THRESH) # int, (nCluster, N)
clusters = proposals_pred[pick_idxs]
cluster_scores = scores_pred[pick_idxs]
cluster_semantic_id = semantic_id[pick_idxs]
nclusters = clusters.shape[0]
##### prepare for evaluation
if cfg.data.eval:
pred_info = {}
pred_info["scene_name"] = test_scene_name
pred_info["conf"] = cluster_scores.cpu().numpy()
pred_info["label_id"] = cluster_semantic_id.cpu().numpy()
pred_info["mask"] = clusters.cpu().numpy()
inst_evaluator.process(inputs, [pred_info])
inference_time = timer.since_last()
##### visual
if cfg.data.visual is not None:
# visual semantic result
gorilla.check_dir(cfg.data.visual)
if cfg.semantic:
pass
# visual instance result
else:
datasets.visualize_instance_mask(
clusters.cpu().numpy(),
test_scene_name,
cfg.data.visual,
os.path.join(data_root, sub_dir),
cluster_scores.cpu().numpy(),
semantic_pred.cpu().numpy(),)
##### save files
if (prepare_flag and cfg.data.save):
f = open(os.path.join(result_dir, test_scene_name + ".txt"), "w")
for proposal_id in range(nclusters):
clusters_i = clusters[proposal_id].cpu().numpy() # [N]
semantic_label = np.argmax(
np.bincount(
semantic_pred[np.where(clusters_i == 1)[0]].cpu()))
score = cluster_scores[proposal_id]
f.write(f"predicted_masks/{test_scene_name}_{proposal_id:03d}.txt "
f"{semantic_label_idx[semantic_label]} {score:.4f}")
if proposal_id < nclusters - 1:
f.write("\n")
content = list(map(lambda x: str(x), clusters_i.tolist()))
content = "\n".join(content)
with open(
os.path.join(
result_dir, "predicted_masks",
test_scene_name + "_%03d.txt" % (proposal_id)),
"w") as cf:
cf.write(content)
# np.savetxt(os.path.join(result_dir, "predicted_masks", test_scene_name + "_%03d.txt" % (proposal_id)), clusters_i, fmt="%d")
f.close()
save_time = timer.since_last()
total_time = timer.since_start()
##### print
if semantic:
logger.info(
f"instance iter: {i + 1}/{len(test_dataloader)} point_num: {N} "
f"time: total {total_time:.2f}s data: {data_time:.2f}s "
f"inference {inference_time:.2f}s save {save_time:.2f}s")
else:
logger.info(
f"instance iter: {i + 1}/{len(test_dataloader)} point_num: {N} "
f"ncluster: {nclusters} time: total {total_time:.2f}s data: {data_time:.2f}s "
f"inference {inference_time:.2f}s save {save_time:.2f}s")
##### evaluation
if cfg.data.eval:
if not semantic:
inst_evaluator.evaluate(prec_rec=False)
evaluator.evaluate()
if __name__ == "__main__":
logger, cfg = init()
##### model
logger.info("=> creating model ...")
logger.info(f"Classes: {cfg.model.classes}")
model = gorilla.build_model(cfg.model)
use_cuda = torch.cuda.is_available()
logger.info(f"cuda available: {use_cuda}")
assert use_cuda
model = model.cuda()
# logger.info(model)
logger.info(f"#classifier parameters (model): {sum([x.nelement() for x in model.parameters()])}")
##### load model
gorilla.load_checkpoint(
model, cfg.pretrain
) # resume from the latest epoch, or specify the epoch to restore
##### evaluate
test(model, cfg, logger)