Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Feature] File I/O migration and reconstruction #2319

Merged
merged 5 commits into from
Mar 13, 2023
Merged

[Feature] File I/O migration and reconstruction #2319

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
4b7ea53
refactor leio
Xiangxu-0103 Mar 1, 2023
a4feb52
update mmengine and mmcv version
Xiangxu-0103 Mar 1, 2023
6df3d09
update
Xiangxu-0103 Mar 2, 2023
9995d0d
update docs
Xiangxu-0103 Mar 6, 2023
a884e49
update version
Xiangxu-0103 Mar 8, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
30 changes: 20 additions & 10 deletions configs/3dssd/3dssd_4xb4_kitti-3d-car.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,6 +9,8 @@
class_names = ['Car']
point_cloud_range = [0, -40, -5, 70, 40, 3]
input_modality = dict(use_lidar=True, use_camera=False)
backend_args = None

db_sampler = dict(
data_root=data_root,
info_path=data_root + 'kitti_dbinfos_train.pkl',
Expand All @@ -17,17 +19,20 @@
classes=class_names,
sample_groups=dict(Car=15),
points_loader=dict(
type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4))

file_client_args = dict(backend='disk')
# Uncomment the following if use ceph or other file clients.
# See https://mmcv.readthedocs.io/en/latest/api.html#mmcv.fileio.FileClient
# for more details.
# file_client_args = dict(
# backend='petrel', path_mapping=dict(data='s3://kitti_data/'))
type='LoadPointsFromFile',
coord_type='LIDAR',
load_dim=4,
use_dim=4,
backend_args=backend_args),
backend_args=backend_args)

train_pipeline = [
dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4),
dict(
type='LoadPointsFromFile',
coord_type='LIDAR',
load_dim=4,
use_dim=4,
backend_args=backend_args),
dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True),
dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
Expand All @@ -52,7 +57,12 @@
]

test_pipeline = [
dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4),
dict(
type='LoadPointsFromFile',
coord_type='LIDAR',
load_dim=4,
use_dim=4,
backend_args=backend_args),
dict(
type='MultiScaleFlipAug3D',
img_scale=(1333, 800),
Expand Down
51 changes: 43 additions & 8 deletions configs/_base_/datasets/kitti-3d-3class.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,6 +6,21 @@
input_modality = dict(use_lidar=True, use_camera=False)
metainfo = dict(classes=class_names)

# Example to use different file client
# Method 1: simply set the data root and let the file I/O module
# automatically Infer from prefix (not support LMDB and Memcache yet)

# data_root = 's3://openmmlab/datasets/detection3d/kitti/'

# Method 2: Use backend_args, file_client_args in versions before 1.1.0rc4
# backend_args = dict(
# backend='petrel',
# path_mapping=dict({
# './data/': 's3://openmmlab/datasets/detection3d/',
# 'data/': 's3://openmmlab/datasets/detection3d/'
# }))
backend_args = None

db_sampler = dict(
data_root=data_root,
info_path=data_root + 'kitti_dbinfos_train.pkl',
Expand All @@ -16,14 +31,20 @@
classes=class_names,
sample_groups=dict(Car=12, Pedestrian=6, Cyclist=6),
points_loader=dict(
type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4))
type='LoadPointsFromFile',
coord_type='LIDAR',
load_dim=4,
use_dim=4,
backend_args=backend_args),
backend_args=backend_args)

train_pipeline = [
dict(
type='LoadPointsFromFile',
coord_type='LIDAR',
load_dim=4, # x, y, z, intensity
use_dim=4),
use_dim=4,
backend_args=backend_args),
dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True),
dict(type='ObjectSample', db_sampler=db_sampler),
dict(
Expand All @@ -45,7 +66,12 @@
keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
]
test_pipeline = [
dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4),
dict(
type='LoadPointsFromFile',
coord_type='LIDAR',
load_dim=4,
use_dim=4,
backend_args=backend_args),
dict(
type='MultiScaleFlipAug3D',
img_scale=(1333, 800),
Expand All @@ -66,7 +92,12 @@
# construct a pipeline for data and gt loading in show function
# please keep its loading function consistent with test_pipeline (e.g. client)
eval_pipeline = [
dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4),
dict(
type='LoadPointsFromFile',
coord_type='LIDAR',
load_dim=4,
use_dim=4,
backend_args=backend_args),
dict(type='Pack3DDetInputs', keys=['points'])
]
train_dataloader = dict(
Expand All @@ -88,7 +119,8 @@
metainfo=metainfo,
# 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')))
box_type_3d='LiDAR',
backend_args=backend_args)))
val_dataloader = dict(
batch_size=1,
num_workers=1,
Expand All @@ -104,7 +136,8 @@
modality=input_modality,
test_mode=True,
metainfo=metainfo,
box_type_3d='LiDAR'))
box_type_3d='LiDAR',
backend_args=backend_args))
test_dataloader = dict(
batch_size=1,
num_workers=1,
Expand All @@ -120,11 +153,13 @@
modality=input_modality,
test_mode=True,
metainfo=metainfo,
box_type_3d='LiDAR'))
box_type_3d='LiDAR',
backend_args=backend_args))
val_evaluator = dict(
type='KittiMetric',
ann_file=data_root + 'kitti_infos_val.pkl',
metric='bbox')
metric='bbox',
backend_args=backend_args)
test_evaluator = val_evaluator

vis_backends = [dict(type='LocalVisBackend')]
Expand Down
51 changes: 43 additions & 8 deletions configs/_base_/datasets/kitti-3d-car.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,6 +6,21 @@
input_modality = dict(use_lidar=True, use_camera=False)
metainfo = dict(classes=class_names)

# Example to use different file client
# Method 1: simply set the data root and let the file I/O module
# automatically Infer from prefix (not support LMDB and Memcache yet)

# data_root = 's3://openmmlab/datasets/detection3d/kitti/'

# Method 2: Use backend_args, file_client_args in versions before 1.1.0rc4
# backend_args = dict(
# backend='petrel',
# path_mapping=dict({
# './data/': 's3://openmmlab/datasets/detection3d/',
# 'data/': 's3://openmmlab/datasets/detection3d/'
# }))
backend_args = None

db_sampler = dict(
data_root=data_root,
info_path=data_root + 'kitti_dbinfos_train.pkl',
Expand All @@ -14,14 +29,20 @@
classes=class_names,
sample_groups=dict(Car=15),
points_loader=dict(
type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4))
type='LoadPointsFromFile',
coord_type='LIDAR',
load_dim=4,
use_dim=4,
backend_args=backend_args),
backend_args=backend_args)

train_pipeline = [
dict(
type='LoadPointsFromFile',
coord_type='LIDAR',
load_dim=4, # x, y, z, intensity
use_dim=4),
use_dim=4,
backend_args=backend_args),
dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True),
dict(type='ObjectSample', db_sampler=db_sampler),
dict(
Expand All @@ -43,7 +64,12 @@
keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
]
test_pipeline = [
dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4),
dict(
type='LoadPointsFromFile',
coord_type='LIDAR',
load_dim=4,
use_dim=4,
backend_args=backend_args),
dict(
type='MultiScaleFlipAug3D',
img_scale=(1333, 800),
Expand All @@ -64,7 +90,12 @@
# construct a pipeline for data and gt loading in show function
# please keep its loading function consistent with test_pipeline (e.g. client)
eval_pipeline = [
dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4),
dict(
type='LoadPointsFromFile',
coord_type='LIDAR',
load_dim=4,
use_dim=4,
backend_args=backend_args),
dict(type='Pack3DDetInputs', keys=['points'])
]
train_dataloader = dict(
Expand All @@ -86,7 +117,8 @@
metainfo=metainfo,
# 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')))
box_type_3d='LiDAR',
backend_args=backend_args)))
val_dataloader = dict(
batch_size=1,
num_workers=1,
Expand All @@ -102,7 +134,8 @@
modality=input_modality,
test_mode=True,
metainfo=metainfo,
box_type_3d='LiDAR'))
box_type_3d='LiDAR',
backend_args=backend_args))
test_dataloader = dict(
batch_size=1,
num_workers=1,
Expand All @@ -118,11 +151,13 @@
modality=input_modality,
test_mode=True,
metainfo=metainfo,
box_type_3d='LiDAR'))
box_type_3d='LiDAR',
backend_args=backend_args))
val_evaluator = dict(
type='KittiMetric',
ann_file=data_root + 'kitti_infos_val.pkl',
metric='bbox')
metric='bbox',
backend_args=backend_args)
test_evaluator = val_evaluator

vis_backends = [dict(type='LocalVisBackend')]
Expand Down
35 changes: 23 additions & 12 deletions configs/_base_/datasets/kitti-mono3d.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,15 +4,23 @@
input_modality = dict(use_lidar=False, use_camera=True)
metainfo = dict(classes=class_names)

file_client_args = dict(backend='disk')
# Uncomment the following if use ceph or other file clients.
# See https://mmcv.readthedocs.io/en/latest/api.html#mmcv.fileio.FileClient
# for more details.
# file_client_args = dict(
# backend='petrel', path_mapping=dict(data='s3://kitti_data/'))
# Example to use different file client
# Method 1: simply set the data root and let the file I/O module
# automatically Infer from prefix (not support LMDB and Memcache yet)

# data_root = 's3://openmmlab/datasets/detection3d/kitti/'

# Method 2: Use backend_args, file_client_args in versions before 1.1.0rc4
# backend_args = dict(
# backend='petrel',
# path_mapping=dict({
# './data/': 's3://openmmlab/datasets/detection3d/',
# 'data/': 's3://openmmlab/datasets/detection3d/'
# }))
backend_args = None

train_pipeline = [
dict(type='LoadImageFromFileMono3D'),
dict(type='LoadImageFromFileMono3D', backend_args=backend_args),
dict(
type='LoadAnnotations3D',
with_bbox=True,
Expand All @@ -31,12 +39,12 @@
]),
]
test_pipeline = [
dict(type='LoadImageFromFileMono3D'),
dict(type='LoadImageFromFileMono3D', backend_args=backend_args),
dict(type='Resize', scale=(1242, 375), keep_ratio=True),
dict(type='Pack3DDetInputs', keys=['img'])
]
eval_pipeline = [
dict(type='LoadImageFromFileMono3D'),
dict(type='LoadImageFromFileMono3D', backend_args=backend_args),
dict(type='Pack3DDetInputs', keys=['img'])
]

Expand All @@ -57,7 +65,8 @@
metainfo=metainfo,
# we use box_type_3d='Camera' in monocular 3d
# detection task
box_type_3d='Camera'))
box_type_3d='Camera',
backend_args=backend_args))
val_dataloader = dict(
batch_size=1,
num_workers=2,
Expand All @@ -74,13 +83,15 @@
load_type='fov_image_based',
metainfo=metainfo,
test_mode=True,
box_type_3d='Camera'))
box_type_3d='Camera',
backend_args=backend_args))
test_dataloader = val_dataloader

val_evaluator = dict(
type='KittiMetric',
ann_file=data_root + 'kitti_infos_val.pkl',
metric='bbox')
metric='bbox',
backend_args=backend_args)

test_evaluator = val_evaluator

Expand Down
Loading