Skip to content

Commit

Permalink
Add config for robust testing
Browse files Browse the repository at this point in the history
  • Loading branch information
@Daniel-xsy
Daniel-xsy committed Feb 22, 2023
1 parent eea49e9 commit 518183b
Show file tree
Hide file tree
Showing 4 changed files with 888 additions and 0 deletions.
284 changes: 284 additions & 0 deletions zoo/PolarFormer/projects/configs/polarformer/polarformer_vovnet_fastbev_aug.py
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,284 @@
_base_ = [
'/nvme/konglingdong/models/mmdetection3d/configs/_base_/datasets/nus-3d.py',
'/nvme/konglingdong/models/mmdetection3d/configs/_base_/default_runtime.py'
]

plugin=True
plugin_dir='projects/mmdet3d_plugin/'

# If point cloud range is changed, the models should also change their point
# cloud range accordingly
point_cloud_range = [-51.2, -51.2, -5.0, 51.2, 51.2, 3.0]
radius_range=[1., 65., 1.] # [start, end, interval]
grid_res = 0.8
voxel_size = [grid_res, grid_res, grid_res]

output_size = [256, 64, 10] # [azimuth, radius, height]
img_norm_cfg = dict(
mean=[103.530, 116.280, 123.675], std=[57.375, 57.120, 58.395], to_rgb=False) # different from r101
# For nuScenes we usually do 10-class detection
class_names = [
'car', 'truck', 'construction_vehicle', 'bus', 'trailer', 'barrier',
'motorcycle', 'bicycle', 'pedestrian', 'traffic_cone'
]

input_modality = dict(
use_lidar=False,
use_camera=True,
use_radar=False,
use_map=False,
use_external=True)

model = dict(
type='PolarFormer',
use_grid_mask=True,
img_backbone=dict(
type='VoVNet',
spec_name='V-99-eSE',
norm_eval=True,
frozen_stages=1,
input_ch=3,
out_features=['stage3', 'stage4', 'stage5']),
img_neck=dict(
type='FPN_TRANS',
num_encoder=0, # encoder is not used here
num_decoder=3,
num_levels=3,
radius_range=radius_range,
use_different_res=True,
use_bev_aug=True,
output_multi_scale=True,
grid_res=grid_res,
pc_range=point_cloud_range,
output_size=output_size,
fpn_cfg=dict(
in_channels=[512, 768, 1024],
out_channels=256,
start_level=1,
add_extra_convs='on_output',
num_outs=3,
relu_before_extra_convs=True),
),
pts_bbox_head=dict(
type='PolarFormerHead',
num_query=900,
num_classes=10,
in_channels=256,
sync_cls_avg_factor=True,
with_box_refine=True,
as_two_stage=False,
radius_range=radius_range,
code_weights = [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.2, 0.2],
transformer=dict(
type='PolarTransformer',
num_feature_levels=3,
encoder=dict(
type='DetrTransformerEncoder',
num_layers=6,
transformerlayers=dict(
type='BaseTransformerLayer',
attn_cfgs=dict(
type='MultiScaleDeformableAttention', embed_dims=256, num_levels=3),
feedforward_channels=1024,
ffn_dropout=0.1,
operation_order=('self_attn', 'norm', 'ffn', 'norm'))),
decoder=dict(
type='PolarTransformerDecoder',
num_layers=6,
return_intermediate=True,
pc_range=point_cloud_range,
radius_range=radius_range,
transformerlayers=dict(
type='DetrTransformerDecoderLayer',
attn_cfgs=[
dict(
type='MultiheadAttention',
embed_dims=256,
num_heads=8,
dropout=0.1),
dict(
type='MultiScaleDeformableAttention',
embed_dims=256,
num_levels=3)
],
feedforward_channels=1024,
ffn_dropout=0.1,
operation_order=('self_attn', 'norm', 'cross_attn', 'norm',
'ffn', 'norm')))),
bbox_coder=dict(
type='NMSFreeCoder',
post_center_range=[-61.2, -61.2, -10.0, 61.2, 61.2, 10.0],
pc_range=point_cloud_range,
max_num=300,
voxel_size=voxel_size,
num_classes=10),
positional_encoding=dict(
type='SinePositionalEncoding',
num_feats=128,
normalize=True,
offset=-0.5),
loss_cls=dict(
type='FocalLoss',
use_sigmoid=True,
gamma=2.0,
alpha=0.25,
loss_weight=2.0),
loss_bbox=dict(type='L1Loss', loss_weight=0.25),
loss_iou=dict(type='GIoULoss', loss_weight=0.0)),
# model training and testing settings
train_cfg=dict(pts=dict(
grid_size=[512, 512, 1],
voxel_size=voxel_size,
point_cloud_range=point_cloud_range,
out_size_factor=4,
assigner=dict(
type='HungarianAssigner3D',
cls_cost=dict(type='FocalLossCost', weight=2.0),
reg_cost=dict(type='BBox3DL1Cost', weight=0.25),
iou_cost=dict(type='IoUCost', weight=0.0), # Fake cost. This is just to make it compatible with DETR head.
pc_range=point_cloud_range))))

dataset_type = 'TransNuScenesDataset'
data_root = '/nvme/share/data/sets/nuScenes/'
corruption_root = '/nvme/konglingdong/data/sets/nuScenes-c/'
anno_root = '../../data/'


file_client_args = dict(backend='disk')

db_sampler = dict(
data_root=data_root,
info_path=data_root + 'nuscenes_dbinfos_train.pkl',
rate=1.0,
prepare=dict(
filter_by_difficulty=[-1],
filter_by_min_points=dict(
car=5,
truck=5,
bus=5,
trailer=5,
construction_vehicle=5,
traffic_cone=5,
barrier=5,
motorcycle=5,
bicycle=5,
pedestrian=5)),
classes=class_names,
sample_groups=dict(
car=2,
truck=3,
construction_vehicle=7,
bus=4,
trailer=6,
barrier=2,
motorcycle=6,
bicycle=6,
pedestrian=2,
traffic_cone=2),
points_loader=dict(
type='LoadPointsFromFile',
coord_type='LIDAR',
load_dim=5,
use_dim=[0, 1, 2, 3, 4],
file_client_args=file_client_args))

train_pipeline = [
dict(type='LoadMultiViewImageFromFiles', to_float32=True),
dict(type='PhotoMetricDistortionMultiViewImage'),
dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True, with_attr_label=False),
dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
dict(type='ObjectNameFilter', classes=class_names),
dict(type='NormalizeMultiviewImage', **img_norm_cfg),
dict(type='PadMultiViewImage', size_divisor=32),
dict(type='DefaultFormatBundle3D', class_names=class_names),
dict(type='Collect3D', keys=['gt_bboxes_3d', 'gt_labels_3d', 'img'],
meta_keys=('filename', 'ori_shape', 'img_shape', 'lidar2img',
'cam2lidar', 'cam_intrinsic',
'depth2img', 'cam2img', 'pad_shape',
'scale_factor', 'flip', 'pcd_horizontal_flip',
'pcd_vertical_flip', 'box_mode_3d', 'box_type_3d',
'img_norm_cfg', 'pcd_trans', 'sample_idx',
'pcd_scale_factor', 'pcd_rotation', 'pts_filename',
'transformation_3d_flow'))
]
test_pipeline = [
dict(type='Custom_LoadMultiViewImageFromFiles', to_float32=True, corruption_root=corruption_root),
dict(type='NormalizeMultiviewImage', **img_norm_cfg),
dict(type='PadMultiViewImage', size_divisor=32),
dict(
type='MultiScaleFlipAug3D',
img_scale=(1333, 800),
pts_scale_ratio=1,
flip=False,
transforms=[
dict(
type='DefaultFormatBundle3D',
class_names=class_names,
with_label=False),
dict(type='Collect3D', keys=['img'],
meta_keys=('filename', 'ori_shape', 'img_shape', 'lidar2img',
'cam2lidar', 'cam_intrinsic',
'depth2img', 'cam2img', 'pad_shape',
'scale_factor', 'flip', 'pcd_horizontal_flip',
'pcd_vertical_flip', 'box_mode_3d', 'box_type_3d',
'img_norm_cfg', 'pcd_trans', 'sample_idx',
'pcd_scale_factor', 'pcd_rotation', 'pts_filename',
'transformation_3d_flow'))
])
]


data = dict(
samples_per_gpu=1,
workers_per_gpu=6,
train=dict(
type=dataset_type,
data_root=data_root,
ann_file=data_root + 'nuscenes_infos_train.pkl',
pipeline=train_pipeline,
classes=class_names,
modality=input_modality,
test_mode=False,
use_valid_flag=True,
# we use box_type_3d='LiDAR' in kitti and nuscenes dataset
# and box_type_3d='Depth' in sunrgbd and scannet dataset.
box_type_3d='LiDAR'),
val=dict(
type=dataset_type,
data_root=data_root,
ann_file=anno_root + 'nuscenes_infos_temporal_val.pkl',
pipeline=test_pipeline,
classes=class_names,
modality=input_modality),
test=dict(
type=dataset_type,
data_root=data_root,
ann_file=anno_root + 'nuscenes_infos_temporal_val.pkl',
pipeline=test_pipeline,
classes=class_names,
modality=input_modality))

optimizer = dict(
type='AdamW',
lr=2e-4,
paramwise_cfg=dict(
custom_keys={
'img_backbone': dict(lr_mult=0.1),
'sampling_offsets': dict(lr_mult=0.1),
'reference_points': dict(lr_mult=0.1),
}),
weight_decay=0.075)
optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2))
# learning policy
lr_config = dict(
policy='CosineAnnealing',
warmup='linear',
warmup_iters=500,
warmup_ratio=1.0 / 3,
min_lr_ratio=1e-3)
total_epochs = 24
evaluation = dict(interval=1, pipeline=test_pipeline)

runner = dict(type='EpochBasedRunner', max_epochs=total_epochs)
load_from='pretrained/dd3_det_final.pth'
find_unused_parameters=True
Loading

0 comments on commit 518183b

Please sign in to comment.