forked from ADLab-AutoDrive/BEVFusion
-
Notifications
You must be signed in to change notification settings - Fork 0
/
fov90_bevf_tf_2x8_1x_nusc.py
executable file
·180 lines (173 loc) · 5.92 KB
/
fov90_bevf_tf_2x8_1x_nusc.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
_base_ = [
'../../_base_/datasets/nusc_fov90_tf.py',
'../../_base_/schedules/schedule_1x.py',
]
point_cloud_range = [-54.0, -54.0, -5.0, 54.0, 54.0, 3.0]
class_names = [
'car', 'truck', 'construction_vehicle', 'bus', 'trailer', 'barrier',
'motorcycle', 'bicycle', 'pedestrian', 'traffic_cone'
]
voxel_size = [0.075, 0.075, 0.2]
out_size_factor = 8
final_dim=(900, 1600) # HxW
downsample=8
imc = 256
input_modality = dict(
use_lidar=True,
use_camera=True,
use_radar=False,
use_map=False,
use_external=False)
num_views = 6
model = dict(
type='BEVF_TransFusion',
se=True,
camera_stream=True,
grid=0.6,
num_views=6,
final_dim=final_dim,
downsample=downsample,
imc=imc,
lic=256 * 2,
lc_fusion=True,
pc_range = point_cloud_range,
img_backbone=dict(
type='CBSwinTransformer',
embed_dim=96,
depths=[2, 2, 6, 2],
num_heads=[3, 6, 12, 24],
window_size=7,
mlp_ratio=4.,
qkv_bias=True,
qk_scale=None,
drop_rate=0.,
attn_drop_rate=0.,
drop_path_rate=0.2,
ape=False,
patch_norm=True,
out_indices=(0, 1, 2, 3),
use_checkpoint=False),
img_neck=dict(
type='FPNC',
final_dim=final_dim,
downsample=downsample,
in_channels=[96, 192, 384, 768],
out_channels=256,
outC=imc,
use_adp=True,
num_outs=5),
pts_voxel_layer=dict(
max_num_points=10,
voxel_size=voxel_size,
max_voxels=(120000, 160000),
point_cloud_range=point_cloud_range),
pts_voxel_encoder=dict(
type='HardSimpleVFE',
num_features=5,
),
pts_middle_encoder=dict(
type='SparseEncoder',
in_channels=5,
sparse_shape=[41, 1440, 1440],
output_channels=128,
order=('conv', 'norm', 'act'),
encoder_channels=((16, 16, 32), (32, 32, 64), (64, 64, 128), (128, 128)),
encoder_paddings=((0, 0, 1), (0, 0, 1), (0, 0, [0, 1, 1]), (0, 0)),
block_type='basicblock'),
pts_backbone=dict(
type='SECOND',
in_channels=256,
out_channels=[128, 256],
layer_nums=[5, 5],
layer_strides=[1, 2],
norm_cfg=dict(type='BN', eps=0.001, momentum=0.01),
conv_cfg=dict(type='Conv2d', bias=False)),
pts_neck=dict(
type='SECONDFPN',
in_channels=[128, 256],
out_channels=[256, 256],
upsample_strides=[1, 2],
norm_cfg=dict(type='BN', eps=0.001, momentum=0.01),
upsample_cfg=dict(type='deconv', bias=False),
use_conv_for_no_stride=True),
pts_bbox_head=dict(
type='TransFusionHead',
fuse_img=False,
num_views=num_views,
in_channels_img=256,
out_size_factor_img=4,
num_proposals=200,
auxiliary=True,
in_channels=256 * 2,
hidden_channel=128,
num_classes=len(class_names),
num_decoder_layers=1,
num_heads=8,
learnable_query_pos=False,
initialize_by_heatmap=True,
nms_kernel_size=3,
ffn_channel=256,
dropout=0.1,
bn_momentum=0.1,
activation='relu',
common_heads=dict(center=(2, 2), height=(1, 2), dim=(3, 2), rot=(2, 2), vel=(2, 2)),
bbox_coder=dict(
type='TransFusionBBoxCoder',
pc_range=point_cloud_range[:2],
voxel_size=voxel_size[:2],
out_size_factor=out_size_factor,
post_center_range=[-61.2, -61.2, -10.0, 61.2, 61.2, 10.0],
score_threshold=0.0,
code_size=10,
),
loss_cls=dict(type='FocalLoss', use_sigmoid=True, gamma=2, alpha=0.25, reduction='mean', loss_weight=1.0),
# loss_iou=dict(type='CrossEntropyLoss', use_sigmoid=True, reduction='mean', loss_weight=0.0),
loss_bbox=dict(type='L1Loss', reduction='mean', loss_weight=0.25),
loss_heatmap=dict(type='GaussianFocalLoss', reduction='mean', loss_weight=1.0),
),
train_cfg=dict(
pts=dict(
dataset='nuScenes',
assigner=dict(
type='HungarianAssigner3D',
iou_calculator=dict(type='BboxOverlaps3D', coordinate='lidar'),
cls_cost=dict(type='FocalLossCost', gamma=2, alpha=0.25, weight=0.15),
reg_cost=dict(type='BBoxBEVL1Cost', weight=0.25),
iou_cost=dict(type='IoU3DCost', weight=0.25)
),
pos_weight=-1,
gaussian_overlap=0.1,
min_radius=2,
grid_size=[1440, 1440, 40], # [x_len, y_len, 1]
voxel_size=voxel_size,
out_size_factor=out_size_factor,
code_weights=[1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.2, 0.2],
point_cloud_range=point_cloud_range)),
test_cfg=dict(
pts=dict(
dataset='nuScenes',
grid_size=[1440, 1440, 40],
out_size_factor=out_size_factor,
pc_range=point_cloud_range[0:2],
voxel_size=voxel_size[:2],
nms_type=None,
)))
optimizer = dict(type='AdamW', lr=0.0001, betas=(0.9, 0.999), weight_decay=0.05,
paramwise_cfg=dict(custom_keys={'absolute_pos_embed': dict(decay_mult=0.),
'relative_position_bias_table': dict(decay_mult=0.),
'norm': dict(decay_mult=0.)}))
optimizer_config = dict(grad_clip=dict(max_norm=10, norm_type=2))
checkpoint_config = dict(interval=1)
log_config = dict(
interval=50,
hooks=[dict(type='TextLoggerHook'),
dict(type='TensorboardLoggerHook')])
dist_params = dict(backend='nccl')
log_level = 'INFO'
load_from = 'work_dirs/bevf_tf_4x8_6e_nusc/epoch_6.pth'
resume_from = None
workflow = [('train', 1)]
gpu_ids = range(0, 8)
data = dict(
samples_per_gpu=2,
workers_per_gpu=6,)