115 lines
3.9 KiB
Python
115 lines
3.9 KiB
Python
# Copyright (c) OpenMMLab. All rights reserved.
|
|
_base_ = [
|
|
'pointpillars_hv_secfpn_kitti.py', 'kitti-3d-3class.py', 'cyclic-40e.py',
|
|
'default_runtime.py'
|
|
]
|
|
|
|
point_cloud_range = [0, -39.68, -3, 69.12, 39.68, 1]
|
|
# dataset settings
|
|
data_root = 'data/kitti/'
|
|
class_names = ['Pedestrian', 'Cyclist', 'Car']
|
|
metainfo = dict(CLASSES=class_names)
|
|
|
|
# PointPillars adopted a different sampling strategies among classes
|
|
db_sampler = dict(
|
|
data_root=data_root,
|
|
info_path=data_root + 'kitti_dbinfos_train.pkl',
|
|
rate=1.0,
|
|
prepare=dict(
|
|
filter_by_difficulty=[-1],
|
|
filter_by_min_points=dict(Car=5, Pedestrian=5, Cyclist=5)),
|
|
classes=class_names,
|
|
sample_groups=dict(Car=15, Pedestrian=15, Cyclist=15),
|
|
points_loader=dict(
|
|
type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4))
|
|
|
|
# PointPillars uses different augmentation hyper parameters
|
|
train_pipeline = [
|
|
dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4),
|
|
dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True),
|
|
dict(type='ObjectSample', db_sampler=db_sampler, use_ground_plane=True),
|
|
dict(type='RandomFlip3D', flip_ratio_bev_horizontal=0.5),
|
|
dict(
|
|
type='GlobalRotScaleTrans',
|
|
rot_range=[-0.78539816, 0.78539816],
|
|
scale_ratio_range=[0.95, 1.05]),
|
|
dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
|
|
dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
|
|
dict(type='PointShuffle'),
|
|
dict(
|
|
type='Pack3DDetInputs',
|
|
keys=['points', 'gt_labels_3d', 'gt_bboxes_3d'])
|
|
]
|
|
test_pipeline = [
|
|
dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4),
|
|
dict(
|
|
type='MultiScaleFlipAug3D',
|
|
img_scale=(1333, 800),
|
|
pts_scale_ratio=1,
|
|
flip=False,
|
|
transforms=[
|
|
dict(
|
|
type='GlobalRotScaleTrans',
|
|
rot_range=[0, 0],
|
|
scale_ratio_range=[1., 1.],
|
|
translation_std=[0, 0, 0]),
|
|
dict(type='RandomFlip3D'),
|
|
dict(
|
|
type='PointsRangeFilter', point_cloud_range=point_cloud_range)
|
|
]),
|
|
dict(type='Pack3DDetInputs', keys=['points'])
|
|
]
|
|
|
|
train_dataloader = dict(
|
|
dataset=dict(dataset=dict(pipeline=train_pipeline, metainfo=metainfo)))
|
|
test_dataloader = dict(dataset=dict(pipeline=test_pipeline, metainfo=metainfo))
|
|
val_dataloader = dict(dataset=dict(pipeline=test_pipeline, metainfo=metainfo))
|
|
# In practice PointPillars also uses a different schedule
|
|
# optimizer
|
|
lr = 0.001
|
|
epoch_num = 80
|
|
optim_wrapper = dict(
|
|
optimizer=dict(lr=lr), clip_grad=dict(max_norm=35, norm_type=2))
|
|
param_scheduler = [
|
|
dict(
|
|
type='CosineAnnealingLR',
|
|
T_max=epoch_num * 0.4,
|
|
eta_min=lr * 10,
|
|
begin=0,
|
|
end=epoch_num * 0.4,
|
|
by_epoch=True,
|
|
convert_to_iter_based=True),
|
|
dict(
|
|
type='CosineAnnealingLR',
|
|
T_max=epoch_num * 0.6,
|
|
eta_min=lr * 1e-4,
|
|
begin=epoch_num * 0.4,
|
|
end=epoch_num * 1,
|
|
by_epoch=True,
|
|
convert_to_iter_based=True),
|
|
dict(
|
|
type='CosineAnnealingMomentum',
|
|
T_max=epoch_num * 0.4,
|
|
eta_min=0.85 / 0.95,
|
|
begin=0,
|
|
end=epoch_num * 0.4,
|
|
by_epoch=True,
|
|
convert_to_iter_based=True),
|
|
dict(
|
|
type='CosineAnnealingMomentum',
|
|
T_max=epoch_num * 0.6,
|
|
eta_min=1,
|
|
begin=epoch_num * 0.4,
|
|
end=epoch_num * 1,
|
|
convert_to_iter_based=True)
|
|
]
|
|
# max_norm=35 is slightly better than 10 for PointPillars in the earlier
|
|
# development of the codebase thus we keep the setting. But we does not
|
|
# specifically tune this parameter.
|
|
# PointPillars usually need longer schedule than second, we simply double
|
|
# the training schedule. Do remind that since we use RepeatDataset and
|
|
# repeat factor is 2, so we actually train 160 epochs.
|
|
train_cfg = dict(by_epoch=True, max_epochs=epoch_num, val_interval=2)
|
|
val_cfg = dict()
|
|
test_cfg = dict()
|