-
Notifications
You must be signed in to change notification settings - Fork 7
/
test_flow_kittisf_benchmark.py
140 lines (119 loc) · 5.48 KB
/
test_flow_kittisf_benchmark.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
import tqdm
import yaml
import argparse
import numpy as np
from collections import OrderedDict
import torch
from utils.pytorch_util import AverageMeter
from metrics.flow_metric import eval_flow
from utils.data_util import upsample_feat
def preproc(pc1, pc2, flow, remove_ground=False, n_sample_point=None):
"""
Follow the same data preprocessing in FlowStep3D.
:param pc1: (N, 3).
:param pc2: (N, 3).
:param flow: (N, 3).
:return:
pc1: (N', 3).
pc2: (N', 3).
flow: (N', 3).
"""
if remove_ground:
is_ground = np.logical_and(pc1[:, 1] < -1.4, pc2[:, 1] < -1.4)
not_ground = np.logical_not(is_ground)
pc1, pc2 = pc1[not_ground], pc2[not_ground]
flow = flow[not_ground]
# Random sampling
if n_sample_point is not None:
indices = pc1.shape[0]
try:
sampled_indices1 = np.random.choice(indices, size=n_sample_point, replace=False, p=None)
sampled_indices2 = np.random.choice(indices, size=n_sample_point, replace=False, p=None)
except:
# replicate some points
sampled_indices1 = np.random.choice(indices, size=n_sample_point, replace=True, p=None)
sampled_indices2 = np.random.choice(indices, size=n_sample_point, replace=True, p=None)
pc1, pc2 = pc1[sampled_indices1], pc2[sampled_indices2]
flow = flow[sampled_indices1]
return pc1, pc2, flow
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('config', type=str, help='Config files')
# Read parameters
args = parser.parse_args()
with open(args.config) as f:
configs = yaml.load(f, Loader=yaml.FullLoader)
for ckey, cvalue in configs.items():
args.__dict__[ckey] = cvalue
# Fix the same random seed as FlowStep3D
np.random.seed(18)
torch.manual_seed(18)
# Setup the network
from models.flownet_kitti import FlowStep3D
flownet = FlowStep3D(npoint=args.flownet['npoint'],
use_instance_norm=args.flownet['use_instance_norm'],
loc_flow_nn=args.flownet['loc_flow_nn'],
loc_flow_rad=args.flownet['loc_flow_rad'],
k_decay_fact=0.5).cuda()
# Load the trained model weights
weight_path = args.save_path
weights_loaded = torch.load(weight_path)['state_dict']
weights = OrderedDict([(k[6:], v) for k, v in weights_loaded.items()])
flownet.load_state_dict(weights)
flownet.eval()
print('Loaded weights from', weight_path)
# Setup the dataset
from datasets.dataset_kittisf import KITTISceneFlowDataset as TestDataset
data_root = args.data['root']
view_sels = [[0, 1], [1, 0]]
mapping_path = 'data_prepare/kittisf/splits/kitti142.txt'
test_set = TestDataset(data_root=data_root,
mapping_path=mapping_path,
downsampled=False,
view_sels=view_sels)
epe_norm_thresh = 0.05
# Setup the dataset to load estimated flow
predflow_path = 'flowstep3d_for-benchmark_R2'
test_set_predflow = TestDataset(data_root=data_root + '_downsampled',
mapping_path=mapping_path,
downsampled=True,
view_sels=view_sels,
predflow_path=predflow_path)
# Iterate over the dataset
eval_meter_fs3d = AverageMeter()
eval_meter = AverageMeter()
n_scenes = len(test_set) // 2
pbar = tqdm.tqdm(total=n_scenes)
for sid in range(n_scenes):
# Only evaluate in forward direction
pcs_org, _, flows_org, _ = test_set[sid * 2]
pcs, _, flow_preds, _ = test_set_predflow[sid * 2]
pc1_org, pc2_org, flow_org = pcs_org[0], pcs_org[1], flows_org[0]
pc, flow_pred = pcs[0], flow_preds[0]
# Preprocess the data before FlowStep3D
pc1_org, pc2_org, flow_org = preproc(pc1_org, pc2_org, flow_org, remove_ground=True, n_sample_point=8192)
# Forward inference with FlowStep3D
flow_org = torch.from_numpy(flow_org).unsqueeze(0)
pc1_org = torch.from_numpy(pc1_org).unsqueeze(0).cuda()
pc2_org = torch.from_numpy(pc2_org).unsqueeze(0).cuda()
flow_preds_fs3d = flownet(pc1_org, pc2_org, pc1_org, pc2_org, iters=5)
flow_pred_fs3d = flow_preds_fs3d[-1]
# Evaluate scene flow estimation from FlowStep3D
epe, acc_strict, acc_relax, outlier = eval_flow(flow_org, flow_pred_fs3d)
eval_meter_fs3d.append_loss({'EPE': epe, 'AccS': acc_strict, 'AccR': acc_relax, 'Outlier': outlier})
# Remove the ground from our downsampled version
is_ground = (pc[:, 1] < -1.4)
not_ground = np.logical_not(is_ground)
pc, flow_pred = pc[not_ground], flow_pred[not_ground]
pc = torch.from_numpy(pc).unsqueeze(0).cuda()
flow_pred = torch.from_numpy(flow_pred).unsqueeze(0).cuda()
# Interpolate from our estimation to obtain flow on specified points
flow_pred = upsample_feat(pc1_org, pc, flow_pred)
# Evaluate scene flow estimation interpolated from ours
epe, acc_strict, acc_relax, outlier = eval_flow(flow_org, flow_pred)
eval_meter.append_loss({'EPE': epe, 'AccS': acc_strict, 'AccR': acc_relax, 'Outlier': outlier})
pbar.update()
eval_avg_fs3d = eval_meter_fs3d.get_mean_loss_dict()
print ('FlowStep3D:', eval_avg_fs3d)
eval_avg = eval_meter.get_mean_loss_dict()
print ('Ours:', eval_avg)