Update waymo dataset, evaluation metrics and related configs for 2.0 (#1663)

* update waymo dataset

* [Fix] Fix all unittests and refactor tests directory and add circle ci in `test-1.x` (#1654)

* add circle ci

* delete github ci

* fix ci

* fix ut

* fix markdown version

* rm

* fix part of uts

* fix comments

* change foler

* refactor test directory

* fix kitti metric ut

* fix all ut
Co-authored-by: VVsssssk <shenkun@pjlab.org.cn>

* add waymo dataset and evaluation metrics

* convert second configs for v2.0

* [Refactor] Unify ceph config (#1677)

* refactor ceph in config

* support metric load ann file from ceph

* add doc string and remove useless code

* [Fix]Fix create data (#1659)

* add circle ci

* delete github ci

* fix ci

* fix ut

* fix markdown version

* rm

* fix part of uts

* fix comments

* change foler

* refactor test directory

* fix kitti metric ut

* fix all ut

* fix creat data
Co-authored-by: ChaimZhu <zhuchenming@pjlab.org.cn>

* [Fix] Fix seg mapping (#1681)

* [Doc]: fix markdown version (#1653)

* [CI] Add circle ci (#1647)

* add circle ci

* delete github ci

* fix ci

* fix ut

* fix markdown version

* rm

* fix seg mapping for scannet

* fix requiremetn

* fix all seg dataet

* resolve commnets
Co-authored-by: VVsssssk <88368822+VVsssssk@users.noreply.github.com>

* [Fix] Fix SSN configs (#1686)

* modify doc string and evaluation file location

* add doc string

* remove path mapping in flieclient args
Co-authored-by: ChaimZhu <zhuchenming@pjlab.org.cn>
Co-authored-by: VVsssssk <shenkun@pjlab.org.cn>
Co-authored-by: VVsssssk <88368822+VVsssssk@users.noreply.github.com>
Co-authored-by: Shilong Zhang <61961338+jshilong@users.noreply.github.com>

configs/3dssd/3dssd_4x4_kitti-3d-car.py

@@ -15,7 +15,9 @@ db_sampler = dict(
     rate=1.0,
     prepare=dict(filter_by_difficulty=[-1], filter_by_min_points=dict(Car=5)),
     classes=class_names,
-    sample_groups=dict(Car=15))
+    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.
@@ -25,17 +27,8 @@ file_client_args = dict(backend='disk')
 #     backend='petrel', path_mapping=dict(data='s3://kitti_data/'))
 
 train_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=4,
-        use_dim=4,
-        file_client_args=file_client_args),
-    dict(
-        type='LoadAnnotations3D',
-        with_bbox_3d=True,
-        with_label_3d=True,
-        file_client_args=file_client_args),
+    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='PointsRangeFilter', point_cloud_range=point_cloud_range),
     dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
     dict(type='ObjectSample', db_sampler=db_sampler),
@@ -59,12 +52,7 @@ train_pipeline = [
 ]
 
 test_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=4,
-        use_dim=4,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4),
     dict(
         type='MultiScaleFlipAug3D',
         img_scale=(1333, 800),

configs/_base_/datasets/kitti-3d-3class.py

@@ -5,6 +5,7 @@ class_names = ['Pedestrian', 'Cyclist', 'Car']
 point_cloud_range = [0, -40, -3, 70.4, 40, 1]
 input_modality = dict(use_lidar=True, use_camera=False)
 metainfo = dict(CLASSES=class_names)
+
 db_sampler = dict(
     data_root=data_root,
     info_path=data_root + 'kitti_dbinfos_train.pkl',
@@ -13,25 +14,13 @@ db_sampler = dict(
         filter_by_difficulty=[-1],
         filter_by_min_points=dict(Car=5, Pedestrian=10, Cyclist=10)),
     classes=class_names,
-    sample_groups=dict(Car=12, Pedestrian=6, Cyclist=6))
-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/'))
+    sample_groups=dict(Car=12, Pedestrian=6, Cyclist=6),
+    points_loader=dict(
+        type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4))
+
 train_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=4,
-        use_dim=4,
-        file_client_args=file_client_args),
-    dict(
-        type='LoadAnnotations3D',
-        with_bbox_3d=True,
-        with_label_3d=True,
-        file_client_args=file_client_args),
+    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),
     dict(
         type='ObjectNoise',
@@ -52,12 +41,7 @@ train_pipeline = [
         keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
 ]
 test_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=4,
-        use_dim=4,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4),
     dict(
         type='MultiScaleFlipAug3D',
         img_scale=(1333, 800),
@@ -78,12 +62,7 @@ test_pipeline = [
 # 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,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4),
     dict(type='Pack3DDetInputs', keys=['points']),
 ]
 train_dataloader = dict(

configs/_base_/datasets/kitti-3d-car.py

@@ -5,31 +5,20 @@ class_names = ['Car']
 point_cloud_range = [0, -40, -3, 70.4, 40, 1]
 input_modality = dict(use_lidar=True, use_camera=False)
 metainfo = dict(CLASSES=class_names)
+
 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)),
     classes=class_names,
-    sample_groups=dict(Car=15))
-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/'))
+    sample_groups=dict(Car=15),
+    points_loader=dict(
+        type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4))
+
 train_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=4,
-        use_dim=4,
-        file_client_args=file_client_args),
-    dict(
-        type='LoadAnnotations3D',
-        with_bbox_3d=True,
-        with_label_3d=True,
-        file_client_args=file_client_args),
+    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),
     dict(
         type='ObjectNoise',
@@ -50,12 +39,7 @@ train_pipeline = [
         keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
 ]
 test_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=4,
-        use_dim=4,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4),
     dict(
         type='MultiScaleFlipAug3D',
         img_scale=(1333, 800),
@@ -76,12 +60,7 @@ test_pipeline = [
 # 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,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4),
     dict(type='Pack3DDetInputs', keys=['points'])
 ]
 train_dataloader = dict(

configs/_base_/datasets/kitti-mono3d.py

@@ -4,21 +4,15 @@ class_names = ['Pedestrian', 'Cyclist', 'Car']
 input_modality = dict(use_lidar=False, use_camera=True)
 metainfo = dict(CLASSES=class_names)
 
-# file_client_args = dict(backend='disk')
+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/kitti/':
-        's3://openmmlab/datasets/detection3d/kitti/',
-        'data/kitti/':
-        's3://openmmlab/datasets/detection3d/kitti/'
-    }))
+# file_client_args = dict(
+#     backend='petrel', path_mapping=dict(data='s3://kitti_data/'))
 
 train_pipeline = [
-    dict(type='LoadImageFromFileMono3D', file_client_args=file_client_args),
+    dict(type='LoadImageFromFileMono3D'),
     dict(
         type='LoadAnnotations3D',
         with_bbox=True,
@@ -37,7 +31,7 @@ train_pipeline = [
         ]),
 ]
 test_pipeline = [
-    dict(type='LoadImageFromFileMono3D', file_client_args=file_client_args),
+    dict(type='LoadImageFromFileMono3D'),
     dict(type='Resize', scale=(1242, 375), keep_ratio=True),
     dict(type='Pack3DDetInputs', keys=['img'])
 ]

configs/_base_/datasets/lyft-3d.py

@@ -12,6 +12,7 @@ data_root = 'data/lyft/'
 # format which requires the information in input_modality.
 input_modality = dict(use_lidar=True, use_camera=False)
 data_prefix = dict(pts='samples/LIDAR_TOP', img='')
+
 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
@@ -22,17 +23,10 @@ file_client_args = dict(backend='disk')
 #         './data/lyft/': 's3://lyft/lyft/',
 #         'data/lyft/': 's3://lyft/lyft/'
 #    }))
+
 train_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=5,
-        use_dim=5,
-        file_client_args=file_client_args),
-    dict(
-        type='LoadPointsFromMultiSweeps',
-        sweeps_num=10,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=5, use_dim=5),
+    dict(type='LoadPointsFromMultiSweeps', sweeps_num=10),
     dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True),
     dict(
         type='GlobalRotScaleTrans',
@@ -48,16 +42,8 @@ train_pipeline = [
         keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
 ]
 test_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=5,
-        use_dim=5,
-        file_client_args=file_client_args),
-    dict(
-        type='LoadPointsFromMultiSweeps',
-        sweeps_num=10,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=5, use_dim=5),
+    dict(type='LoadPointsFromMultiSweeps', sweeps_num=10),
     dict(
         type='MultiScaleFlipAug3D',
         img_scale=(1333, 800),
@@ -78,16 +64,8 @@ test_pipeline = [
 # 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=5,
-        use_dim=5,
-        file_client_args=file_client_args),
-    dict(
-        type='LoadPointsFromMultiSweeps',
-        sweeps_num=10,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=5, use_dim=5),
+    dict(type='LoadPointsFromMultiSweeps', sweeps_num=10),
     dict(type='Pack3DDetInputs', keys=['points'])
 ]
 train_dataloader = dict(

configs/_base_/datasets/nuim_instance.py

@@ -6,6 +6,14 @@ class_names = [
 ]
 img_norm_cfg = dict(
     mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
+
+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://nuimages'))
+
 train_pipeline = [
     dict(type='LoadImageFromFile'),
     dict(type='LoadAnnotations', with_bbox=True, with_mask=True),

configs/_base_/datasets/nus-3d.py

@@ -12,8 +12,9 @@ data_root = 'data/nuscenes/'
 # Input modality for nuScenes dataset, this is consistent with the submission
 # format which requires the information in input_modality.
 input_modality = dict(use_lidar=True, use_camera=False)
-file_client_args = dict(backend='disk')
 data_prefix = dict(pts='samples/LIDAR_TOP', img='')
+
+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.
@@ -23,17 +24,10 @@ data_prefix = dict(pts='samples/LIDAR_TOP', img='')
 #         './data/nuscenes/': 's3://nuscenes/nuscenes/',
 #         'data/nuscenes/': 's3://nuscenes/nuscenes/'
 #     }))
+
 train_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=5,
-        use_dim=5,
-        file_client_args=file_client_args),
-    dict(
-        type='LoadPointsFromMultiSweeps',
-        sweeps_num=10,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=5, use_dim=5),
+    dict(type='LoadPointsFromMultiSweeps', sweeps_num=10),
     dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True),
     dict(
         type='GlobalRotScaleTrans',
@@ -50,17 +44,8 @@ train_pipeline = [
         keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
 ]
 test_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=5,
-        use_dim=5,
-        file_client_args=file_client_args),
-    dict(
-        type='LoadPointsFromMultiSweeps',
-        sweeps_num=10,
-        test_mode=True,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=5, use_dim=5),
+    dict(type='LoadPointsFromMultiSweeps', sweeps_num=10, test_mode=True),
     dict(
         type='MultiScaleFlipAug3D',
         img_scale=(1333, 800),
@@ -81,17 +66,8 @@ test_pipeline = [
 # 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=5,
-        use_dim=5,
-        file_client_args=file_client_args),
-    dict(
-        type='LoadPointsFromMultiSweeps',
-        sweeps_num=10,
-        test_mode=True,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=5, use_dim=5),
+    dict(type='LoadPointsFromMultiSweeps', sweeps_num=10, test_mode=True),
     dict(type='Pack3DDetInputs', keys=['points'])
 ]
 train_dataloader = dict(

configs/_base_/datasets/nus-mono3d.py

@@ -9,21 +9,21 @@ metainfo = dict(CLASSES=class_names)
 # format which requires the information in input_modality.
 input_modality = dict(use_lidar=False, use_camera=True)
 
-# file_client_args = dict(backend='disk')
+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/nuscenes/':
-        's3://openmmlab/datasets/detection3d/nuscenes/',
-        'data/nuscenes/':
-        's3://openmmlab/datasets/detection3d/nuscenes/'
-    }))
+# file_client_args = dict(
+#     backend='petrel',
+#     path_mapping=dict({
+#         './data/nuscenes/':
+#         's3://openmmlab/datasets/detection3d/nuscenes/',
+#         'data/nuscenes/':
+#         's3://openmmlab/datasets/detection3d/nuscenes/'
+#     }))
 
 train_pipeline = [
-    dict(type='LoadImageFromFileMono3D', file_client_args=file_client_args),
+    dict(type='LoadImageFromFileMono3D'),
     dict(
         type='LoadAnnotations3D',
         with_bbox=True,
@@ -32,7 +32,7 @@ train_pipeline = [
         with_bbox_3d=True,
         with_label_3d=True,
         with_bbox_depth=True),
-    dict(type='Resize', img_scale=(1600, 900), keep_ratio=True),
+    dict(type='Resize', scale=(1600, 900), keep_ratio=True),
     dict(type='RandomFlip3D', flip_ratio_bev_horizontal=0.5),
     dict(
         type='Pack3DDetInputs',
@@ -43,7 +43,7 @@ train_pipeline = [
 ]
 
 test_pipeline = [
-    dict(type='LoadImageFromFileMono3D', file_client_args=file_client_args),
+    dict(type='LoadImageFromFileMono3D'),
     dict(type='mmdet.Resize', scale=(1600, 900), keep_ratio=True),
     dict(type='Pack3DDetInputs', keys=['img'])
 ]

configs/_base_/datasets/range100_lyft-3d.py

@@ -27,16 +27,8 @@ file_client_args = dict(backend='disk')
 #         'data/lyft/': 's3://lyft/lyft/'
 #    }))
 train_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=5,
-        use_dim=5,
-        file_client_args=file_client_args),
-    dict(
-        type='LoadPointsFromMultiSweeps',
-        sweeps_num=10,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=5, use_dim=5),
+    dict(type='LoadPointsFromMultiSweeps', sweeps_num=10),
     dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True),
     dict(
         type='GlobalRotScaleTrans',
@@ -51,16 +43,8 @@ train_pipeline = [
     dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
 ]
 test_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=5,
-        use_dim=5,
-        file_client_args=file_client_args),
-    dict(
-        type='LoadPointsFromMultiSweeps',
-        sweeps_num=10,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=5, use_dim=5),
+    dict(type='LoadPointsFromMultiSweeps', sweeps_num=10),
     dict(
         type='MultiScaleFlipAug3D',
         img_scale=(1333, 800),
@@ -85,16 +69,8 @@ test_pipeline = [
 # 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=5,
-        use_dim=5,
-        file_client_args=file_client_args),
-    dict(
-        type='LoadPointsFromMultiSweeps',
-        sweeps_num=10,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=5, use_dim=5),
+    dict(type='LoadPointsFromMultiSweeps', sweeps_num=10),
     dict(
         type='DefaultFormatBundle3D',
         class_names=class_names,

configs/_base_/datasets/s3dis-3d-5class.py

@@ -6,6 +6,17 @@ class_names = ('table', 'chair', 'sofa', 'bookcase', 'board')
 train_area = [1, 2, 3, 4, 6]
 test_area = 5
 
+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/s3dis/':
+#         's3://s3dis/',
+#     }))
+
 train_pipeline = [
     dict(
         type='LoadPointsFromFile',

configs/_base_/datasets/s3dis_seg-3d-13class.py

@@ -18,9 +18,7 @@ file_client_args = dict(backend='disk')
 #     backend='petrel',
 #     path_mapping=dict({
 #         './data/s3dis/':
-#         's3://openmmlab/datasets/detection3d/s3dis_processed/',
-#         'data/s3dis/':
-#         's3://openmmlab/datasets/detection3d/s3dis_processed/'
+#         's3://s3dis/',
 #     }))
 
 num_points = 4096
@@ -33,15 +31,13 @@ train_pipeline = [
         shift_height=False,
         use_color=True,
         load_dim=6,
-        use_dim=[0, 1, 2, 3, 4, 5],
-        file_client_args=file_client_args),
+        use_dim=[0, 1, 2, 3, 4, 5]),
     dict(
         type='LoadAnnotations3D',
         with_bbox_3d=False,
         with_label_3d=False,
         with_mask_3d=False,
-        with_seg_3d=True,
-        file_client_args=file_client_args),
+        with_seg_3d=True),
     dict(type='PointSegClassMapping'),
     dict(
         type='IndoorPatchPointSample',
@@ -61,8 +57,7 @@ test_pipeline = [
         shift_height=False,
         use_color=True,
         load_dim=6,
-        use_dim=[0, 1, 2, 3, 4, 5],
-        file_client_args=file_client_args),
+        use_dim=[0, 1, 2, 3, 4, 5]),
     dict(type='NormalizePointsColor', color_mean=None),
     dict(
         # a wrapper in order to successfully call test function
@@ -95,8 +90,7 @@ eval_pipeline = [
         shift_height=False,
         use_color=True,
         load_dim=6,
-        use_dim=[0, 1, 2, 3, 4, 5],
-        file_client_args=file_client_args),
+        use_dim=[0, 1, 2, 3, 4, 5]),
     dict(type='NormalizePointsColor', color_mean=None),
     dict(type='Pack3DDetInputs', keys=['points'])
 ]

configs/_base_/datasets/scannet-3d-18class.py

@@ -7,6 +7,18 @@ metainfo = dict(
              'bookshelf', 'picture', 'counter', 'desk', 'curtain',
              'refrigerator', 'showercurtrain', 'toilet', 'sink', 'bathtub',
              'garbagebin'))
+
+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/scannet/':
+#         's3://scannet/',
+#     }))
+
 train_pipeline = [
     dict(
         type='LoadPointsFromFile',
@@ -21,11 +33,7 @@ train_pipeline = [
         with_mask_3d=True,
         with_seg_3d=True),
     dict(type='GlobalAlignment', rotation_axis=2),
-    dict(
-        type='PointSegClassMapping',
-        valid_cat_ids=(3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 16, 24, 28, 33, 34,
-                       36, 39),
-        max_cat_id=40),
+    dict(type='PointSegClassMapping'),
     dict(type='PointSample', num_points=40000),
     dict(
         type='RandomFlip3D',

configs/_base_/datasets/scannet_seg-3d-20class.py

@@ -20,9 +20,7 @@ file_client_args = dict(backend='disk')
 #     backend='petrel',
 #     path_mapping=dict({
 #         './data/scannet/':
-#         's3://openmmlab/datasets/detection3d/scannet_processed/',
-#         'data/scannet/':
-#         's3://openmmlab/datasets/detection3d/scannet_processed/'
+#         's3://scannet/',
 #     }))
 
 num_points = 8192
@@ -33,15 +31,13 @@ train_pipeline = [
         shift_height=False,
         use_color=True,
         load_dim=6,
-        use_dim=[0, 1, 2, 3, 4, 5],
-        file_client_args=file_client_args),
+        use_dim=[0, 1, 2, 3, 4, 5]),
     dict(
         type='LoadAnnotations3D',
         with_bbox_3d=False,
         with_label_3d=False,
         with_mask_3d=False,
-        with_seg_3d=True,
-        file_client_args=file_client_args),
+        with_seg_3d=True),
     dict(type='PointSegClassMapping'),
     dict(
         type='IndoorPatchPointSample',
@@ -61,8 +57,7 @@ test_pipeline = [
         shift_height=False,
         use_color=True,
         load_dim=6,
-        use_dim=[0, 1, 2, 3, 4, 5],
-        file_client_args=file_client_args),
+        use_dim=[0, 1, 2, 3, 4, 5]),
     dict(type='NormalizePointsColor', color_mean=None),
     dict(
         # a wrapper in order to successfully call test function
@@ -95,8 +90,7 @@ eval_pipeline = [
         shift_height=False,
         use_color=True,
         load_dim=6,
-        use_dim=[0, 1, 2, 3, 4, 5],
-        file_client_args=file_client_args),
+        use_dim=[0, 1, 2, 3, 4, 5]),
     dict(type='NormalizePointsColor', color_mean=None),
     dict(type='Pack3DDetInputs', keys=['points'])
 ]

configs/_base_/datasets/sunrgbd-3d-10class.py

@@ -5,6 +5,17 @@ class_names = ('bed', 'table', 'sofa', 'chair', 'toilet', 'desk', 'dresser',
 
 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/sunrgbd/':
+#         's3://sunrgbd/',
+#     }))
+
 train_pipeline = [
     dict(
         type='LoadPointsFromFile',

configs/_base_/datasets/waymoD5-3d-3class.py

@@ -11,6 +11,8 @@ file_client_args = dict(backend='disk')
 #     backend='petrel', path_mapping=dict(data='s3://waymo_data/'))
 
 class_names = ['Car', 'Pedestrian', 'Cyclist']
+metainfo = dict(CLASSES=class_names)
+
 point_cloud_range = [-74.88, -74.88, -2, 74.88, 74.88, 4]
 input_modality = dict(use_lidar=True, use_camera=False)
 db_sampler = dict(
@@ -26,21 +28,11 @@ db_sampler = dict(
         type='LoadPointsFromFile',
         coord_type='LIDAR',
         load_dim=6,
-        use_dim=[0, 1, 2, 3, 4],
-        file_client_args=file_client_args))
+        use_dim=[0, 1, 2, 3, 4]))
 
 train_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=6,
-        use_dim=5,
-        file_client_args=file_client_args),
-    dict(
-        type='LoadAnnotations3D',
-        with_bbox_3d=True,
-        with_label_3d=True,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=6, use_dim=5),
+    dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True),
     dict(type='ObjectSample', db_sampler=db_sampler),
     dict(
         type='RandomFlip3D',
@@ -54,16 +46,12 @@ train_pipeline = [
     dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
     dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
     dict(type='PointShuffle'),
-    dict(type='DefaultFormatBundle3D', class_names=class_names),
-    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
+    dict(
+        type='Pack3DDetInputs',
+        keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
 ]
 test_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=6,
-        use_dim=5,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=6, use_dim=5),
     dict(
         type='MultiScaleFlipAug3D',
         img_scale=(1333, 800),
@@ -77,13 +65,9 @@ test_pipeline = [
                 translation_std=[0, 0, 0]),
             dict(type='RandomFlip3D'),
             dict(
-                type='PointsRangeFilter', point_cloud_range=point_cloud_range),
-            dict(
-                type='DefaultFormatBundle3D',
-                class_names=class_names,
-                with_label=False),
-            dict(type='Collect3D', keys=['points'])
-        ])
+                type='PointsRangeFilter', point_cloud_range=point_cloud_range)
+        ]),
+    dict(type='Pack3DDetInputs', keys=['points'])
 ]
 # construct a pipeline for data and gt loading in show function
 # please keep its loading function consistent with test_pipeline (e.g. client)
@@ -94,52 +78,69 @@ eval_pipeline = [
         load_dim=6,
         use_dim=5,
         file_client_args=file_client_args),
-    dict(
-        type='DefaultFormatBundle3D',
-        class_names=class_names,
-        with_label=False),
-    dict(type='Collect3D', keys=['points'])
+    dict(type='Pack3DDetInputs', keys=['points']),
 ]
 
-data = dict(
-    samples_per_gpu=2,
-    workers_per_gpu=4,
-    train=dict(
+train_dataloader = dict(
+    batch_size=2,
+    num_workers=2,
+    persistent_workers=True,
+    sampler=dict(type='DefaultSampler', shuffle=True),
+    dataset=dict(
         type='RepeatDataset',
         times=2,
         dataset=dict(
             type=dataset_type,
             data_root=data_root,
-            ann_file=data_root + 'waymo_infos_train.pkl',
-            split='training',
+            ann_file='waymo_infos_train.pkl',
+            data_prefix=dict(pts='training/velodyne'),
             pipeline=train_pipeline,
             modality=input_modality,
-            classes=class_names,
             test_mode=False,
+            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',
             # load one frame every five frames
-            load_interval=5)),
-    val=dict(
+            load_interval=5)))
+val_dataloader = dict(
+    batch_size=1,
+    num_workers=1,
+    persistent_workers=True,
+    drop_last=False,
+    sampler=dict(type='DefaultSampler', shuffle=False),
+    dataset=dict(
         type=dataset_type,
         data_root=data_root,
-        ann_file=data_root + 'waymo_infos_val.pkl',
-        split='training',
-        pipeline=test_pipeline,
+        data_prefix=dict(pts='training/velodyne'),
+        ann_file='waymo_infos_val.pkl',
+        pipeline=eval_pipeline,
         modality=input_modality,
-        classes=class_names,
         test_mode=True,
-        box_type_3d='LiDAR'),
-    test=dict(
+        metainfo=metainfo,
+        box_type_3d='LiDAR'))
+
+test_dataloader = dict(
+    batch_size=1,
+    num_workers=1,
+    persistent_workers=True,
+    drop_last=False,
+    sampler=dict(type='DefaultSampler', shuffle=False),
+    dataset=dict(
         type=dataset_type,
         data_root=data_root,
-        ann_file=data_root + 'waymo_infos_val.pkl',
-        split='training',
-        pipeline=test_pipeline,
+        data_prefix=dict(pts='training/velodyne'),
+        ann_file='waymo_infos_val.pkl',
+        pipeline=eval_pipeline,
         modality=input_modality,
-        classes=class_names,
         test_mode=True,
+        metainfo=metainfo,
         box_type_3d='LiDAR'))
 
-evaluation = dict(interval=24, pipeline=eval_pipeline)
+val_evaluator = dict(
+    type='WaymoMetric',
+    ann_file='./data/waymo/kitti_format/waymo_infos_val.pkl',
+    waymo_bin_file='./data/waymo/waymo_format/gt.bin',
+    data_root='./data/waymo/waymo_format',
+    file_client_args=file_client_args)
+test_evaluator = val_evaluator

configs/_base_/datasets/waymoD5-3d-car.py

@@ -11,6 +11,8 @@ file_client_args = dict(backend='disk')
 #     backend='petrel', path_mapping=dict(data='s3://waymo_data/'))
 
 class_names = ['Car']
+metainfo = dict(CLASSES=class_names)
+
 point_cloud_range = [-74.88, -74.88, -2, 74.88, 74.88, 4]
 input_modality = dict(use_lidar=True, use_camera=False)
 db_sampler = dict(
@@ -24,21 +26,11 @@ db_sampler = dict(
         type='LoadPointsFromFile',
         coord_type='LIDAR',
         load_dim=6,
-        use_dim=[0, 1, 2, 3, 4],
-        file_client_args=file_client_args))
+        use_dim=[0, 1, 2, 3, 4]))
 
 train_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=6,
-        use_dim=5,
-        file_client_args=file_client_args),
-    dict(
-        type='LoadAnnotations3D',
-        with_bbox_3d=True,
-        with_label_3d=True,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=6, use_dim=5),
+    dict(type='LoadAnnotations3D', with_bbox_3d=True, with_label_3d=True),
     dict(type='ObjectSample', db_sampler=db_sampler),
     dict(
         type='RandomFlip3D',
@@ -52,16 +44,12 @@ train_pipeline = [
     dict(type='PointsRangeFilter', point_cloud_range=point_cloud_range),
     dict(type='ObjectRangeFilter', point_cloud_range=point_cloud_range),
     dict(type='PointShuffle'),
-    dict(type='DefaultFormatBundle3D', class_names=class_names),
-    dict(type='Collect3D', keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
+    dict(
+        type='Pack3DDetInputs',
+        keys=['points', 'gt_bboxes_3d', 'gt_labels_3d'])
 ]
 test_pipeline = [
-    dict(
-        type='LoadPointsFromFile',
-        coord_type='LIDAR',
-        load_dim=6,
-        use_dim=5,
-        file_client_args=file_client_args),
+    dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=6, use_dim=5),
     dict(
         type='MultiScaleFlipAug3D',
         img_scale=(1333, 800),
@@ -75,13 +63,9 @@ test_pipeline = [
                 translation_std=[0, 0, 0]),
             dict(type='RandomFlip3D'),
             dict(
-                type='PointsRangeFilter', point_cloud_range=point_cloud_range),
-            dict(
-                type='DefaultFormatBundle3D',
-                class_names=class_names,
-                with_label=False),
-            dict(type='Collect3D', keys=['points'])
-        ])
+                type='PointsRangeFilter', point_cloud_range=point_cloud_range)
+        ]),
+    dict(type='Pack3DDetInputs', keys=['points'])
 ]
 # construct a pipeline for data and gt loading in show function
 # please keep its loading function consistent with test_pipeline (e.g. client)
@@ -92,52 +76,69 @@ eval_pipeline = [
         load_dim=6,
         use_dim=5,
         file_client_args=file_client_args),
-    dict(
-        type='DefaultFormatBundle3D',
-        class_names=class_names,
-        with_label=False),
-    dict(type='Collect3D', keys=['points'])
+    dict(type='Pack3DDetInputs', keys=['points']),
 ]
 
-data = dict(
-    samples_per_gpu=2,
-    workers_per_gpu=4,
-    train=dict(
+train_dataloader = dict(
+    batch_size=2,
+    num_workers=2,
+    persistent_workers=True,
+    sampler=dict(type='DefaultSampler', shuffle=True),
+    dataset=dict(
         type='RepeatDataset',
         times=2,
         dataset=dict(
             type=dataset_type,
             data_root=data_root,
-            ann_file=data_root + 'waymo_infos_train.pkl',
-            split='training',
+            ann_file='waymo_infos_train.pkl',
+            data_prefix=dict(pts='training/velodyne'),
             pipeline=train_pipeline,
             modality=input_modality,
-            classes=class_names,
             test_mode=False,
+            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',
             # load one frame every five frames
-            load_interval=5)),
-    val=dict(
+            load_interval=5)))
+val_dataloader = dict(
+    batch_size=1,
+    num_workers=1,
+    persistent_workers=True,
+    drop_last=False,
+    sampler=dict(type='DefaultSampler', shuffle=False),
+    dataset=dict(
         type=dataset_type,
         data_root=data_root,
-        ann_file=data_root + 'waymo_infos_val.pkl',
-        split='training',
-        pipeline=test_pipeline,
+        data_prefix=dict(pts='training/velodyne'),
+        ann_file='waymo_infos_val.pkl',
+        pipeline=eval_pipeline,
         modality=input_modality,
-        classes=class_names,
         test_mode=True,
-        box_type_3d='LiDAR'),
-    test=dict(
+        metainfo=metainfo,
+        box_type_3d='LiDAR'))
+
+test_dataloader = dict(
+    batch_size=1,
+    num_workers=1,
+    persistent_workers=True,
+    drop_last=False,
+    sampler=dict(type='DefaultSampler', shuffle=False),
+    dataset=dict(
         type=dataset_type,
         data_root=data_root,
-        ann_file=data_root + 'waymo_infos_val.pkl',
-        split='training',
-        pipeline=test_pipeline,
+        data_prefix=dict(pts='training/velodyne'),
+        ann_file='waymo_infos_val.pkl',
+        pipeline=eval_pipeline,
         modality=input_modality,
-        classes=class_names,
         test_mode=True,
+        metainfo=metainfo,
         box_type_3d='LiDAR'))
 
-evaluation = dict(interval=24, pipeline=eval_pipeline)
+val_evaluator = dict(
+    type='WaymoMetric',
+    ann_file='./data/waymo/kitti_format/waymo_infos_val.pkl',
+    waymo_bin_file='./data/waymo/waymo_format/gt.bin',
+    data_root='./data/waymo/waymo_format',
+    file_client_args=file_client_args)
+test_evaluator = val_evaluator

configs/_base_/models/hv_pointpillars_secfpn_waymo.py

@@ -6,6 +6,7 @@
 voxel_size = [0.32, 0.32, 6]
 model = dict(
     type='MVXFasterRCNN',
+    data_preprocessor=dict(type='Det3DDataPreprocessor'),
     pts_voxel_layer=dict(
         max_num_points=20,
         point_cloud_range=[-74.88, -74.88, -2, 74.88, 74.88, 4],
@@ -58,34 +59,36 @@ model = dict(
         dir_offset=-0.7854,  # -pi / 4
         bbox_coder=dict(type='DeltaXYZWLHRBBoxCoder', code_size=7),
         loss_cls=dict(
-            type='FocalLoss',
+            type='mmdet.FocalLoss',
             use_sigmoid=True,
             gamma=2.0,
             alpha=0.25,
             loss_weight=1.0),
-        loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0),
+        loss_bbox=dict(
+            type='mmdet.SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0),
         loss_dir=dict(
-            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2)),
+            type='mmdet.CrossEntropyLoss', use_sigmoid=False,
+            loss_weight=0.2)),
     # model training and testing settings
     train_cfg=dict(
         pts=dict(
             assigner=[
                 dict(  # car
-                    type='MaxIoUAssigner',
+                    type='Max3DIoUAssigner',
                     iou_calculator=dict(type='BboxOverlapsNearest3D'),
                     pos_iou_thr=0.55,
                     neg_iou_thr=0.4,
                     min_pos_iou=0.4,
                     ignore_iof_thr=-1),
                 dict(  # cyclist
-                    type='MaxIoUAssigner',
+                    type='Max3DIoUAssigner',
                     iou_calculator=dict(type='BboxOverlapsNearest3D'),
                     pos_iou_thr=0.5,
                     neg_iou_thr=0.3,
                     min_pos_iou=0.3,
                     ignore_iof_thr=-1),
                 dict(  # pedestrian
-                    type='MaxIoUAssigner',
+                    type='Max3DIoUAssigner',
                     iou_calculator=dict(type='BboxOverlapsNearest3D'),
                     pos_iou_thr=0.5,
                     neg_iou_thr=0.3,

configs/_base_/models/hv_second_secfpn_waymo.py

@@ -5,32 +5,33 @@
 # keys in the config.
 voxel_size = [0.08, 0.08, 0.1]
 model = dict(
-    type='VoxelNet',
-    voxel_layer=dict(
-        max_num_points=10,
+    type='MVXFasterRCNN',
+    data_preprocessor=dict(type='Det3DDataPreprocessor'),
+    pts_voxel_layer=dict(
+        max_num_points=20,
         point_cloud_range=[-76.8, -51.2, -2, 76.8, 51.2, 4],
         voxel_size=voxel_size,
         max_voxels=(80000, 90000)),
-    voxel_encoder=dict(type='HardSimpleVFE', num_features=5),
-    middle_encoder=dict(
+    pts_voxel_encoder=dict(type='HardSimpleVFE', num_features=5),
+    pts_middle_encoder=dict(
         type='SparseEncoder',
         in_channels=5,
         sparse_shape=[61, 1280, 1920],
         order=('conv', 'norm', 'act')),
-    backbone=dict(
+    pts_backbone=dict(
         type='SECOND',
         in_channels=384,
         norm_cfg=dict(type='naiveSyncBN2d', eps=1e-3, momentum=0.01),
         layer_nums=[5, 5],
         layer_strides=[1, 2],
         out_channels=[128, 256]),
-    neck=dict(
+    pts_neck=dict(
         type='SECONDFPN',
         norm_cfg=dict(type='naiveSyncBN2d', eps=1e-3, momentum=0.01),
         in_channels=[128, 256],
         upsample_strides=[1, 2],
         out_channels=[256, 256]),
-    bbox_head=dict(
+    pts_bbox_head=dict(
         type='Anchor3DHead',
         num_classes=3,
         in_channels=512,
@@ -38,13 +39,15 @@ model = dict(
         use_direction_classifier=True,
         anchor_generator=dict(
             type='AlignedAnchor3DRangeGenerator',
-            ranges=[[-76.8, -51.2, -0.0345, 76.8, 51.2, -0.0345],
-                    [-76.8, -51.2, 0, 76.8, 51.2, 0],
-                    [-76.8, -51.2, -0.1188, 76.8, 51.2, -0.1188]],
+            ranges=[
+                [-76.8, -51.2, -0.0345, 76.8, 51.2, -0.0345],
+                [-76.8, -51.2, -0.1188, 76.8, 51.2, -0.1188],
+                [-76.8, -51.2, 0, 76.8, 51.2, 0],
+            ],
             sizes=[
                 [4.73, 2.08, 1.77],  # car
-                [0.91, 0.84, 1.74],  # pedestrian
-                [1.81, 0.84, 1.77]  # cyclist
+                [1.81, 0.84, 1.77],  # pedestrian
+                [0.91, 0.84, 1.74],  # cyclist
             ],
             rotations=[0, 1.57],
             reshape_out=False),
@@ -52,48 +55,52 @@ model = dict(
         dir_offset=-0.7854,  # -pi / 4
         bbox_coder=dict(type='DeltaXYZWLHRBBoxCoder', code_size=7),
         loss_cls=dict(
-            type='FocalLoss',
+            type='mmdet.FocalLoss',
             use_sigmoid=True,
             gamma=2.0,
             alpha=0.25,
             loss_weight=1.0),
-        loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0),
+        loss_bbox=dict(
+            type='mmdet.SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0),
         loss_dir=dict(
-            type='CrossEntropyLoss', use_sigmoid=False, loss_weight=0.2)),
+            type='mmdet.CrossEntropyLoss', use_sigmoid=False,
+            loss_weight=0.2)),
     # model training and testing settings
     train_cfg=dict(
-        assigner=[
-            dict(  # car
-                type='MaxIoUAssigner',
-                iou_calculator=dict(type='BboxOverlapsNearest3D'),
-                pos_iou_thr=0.55,
-                neg_iou_thr=0.4,
-                min_pos_iou=0.4,
-                ignore_iof_thr=-1),
-            dict(  # pedestrian
-                type='MaxIoUAssigner',
-                iou_calculator=dict(type='BboxOverlapsNearest3D'),
-                pos_iou_thr=0.5,
-                neg_iou_thr=0.3,
-                min_pos_iou=0.3,
-                ignore_iof_thr=-1),
-            dict(  # cyclist
-                type='MaxIoUAssigner',
-                iou_calculator=dict(type='BboxOverlapsNearest3D'),
-                pos_iou_thr=0.5,
-                neg_iou_thr=0.3,
-                min_pos_iou=0.3,
-                ignore_iof_thr=-1)
-        ],
-        allowed_border=0,
-        code_weight=[1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],
-        pos_weight=-1,
-        debug=False),
+        pts=dict(
+            assigner=[
+                dict(  # car
+                    type='Max3DIoUAssigner',
+                    iou_calculator=dict(type='BboxOverlapsNearest3D'),
+                    pos_iou_thr=0.55,
+                    neg_iou_thr=0.4,
+                    min_pos_iou=0.4,
+                    ignore_iof_thr=-1),
+                dict(  # pedestrian
+                    type='Max3DIoUAssigner',
+                    iou_calculator=dict(type='BboxOverlapsNearest3D'),
+                    pos_iou_thr=0.5,
+                    neg_iou_thr=0.3,
+                    min_pos_iou=0.3,
+                    ignore_iof_thr=-1),
+                dict(  # cyclist
+                    type='Max3DIoUAssigner',
+                    iou_calculator=dict(type='BboxOverlapsNearest3D'),
+                    pos_iou_thr=0.5,
+                    neg_iou_thr=0.3,
+                    min_pos_iou=0.3,
+                    ignore_iof_thr=-1),
+            ],
+            allowed_border=0,
+            code_weight=[1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],
+            pos_weight=-1,
+            debug=False)),
     test_cfg=dict(
-        use_rotate_nms=True,
-        nms_across_levels=False,
-        nms_pre=4096,
-        nms_thr=0.25,
-        score_thr=0.1,
-        min_bbox_size=0,
-        max_num=500))
+        pts=dict(
+            use_rotate_nms=True,
+            nms_across_levels=False,
+            nms_pre=4096,
+            nms_thr=0.25,
+            score_thr=0.1,
+            min_bbox_size=0,
+            max_num=500)))

configs/benchmark/hv_PartA2_secfpn_4x8_cyclic_80e_pcdet_kitti-3d-3class.py

@@ -213,7 +213,9 @@ db_sampler = dict(
         filter_by_difficulty=[-1],
         filter_by_min_points=dict(Car=5, Pedestrian=5, Cyclist=5)),
     classes=class_names,
-    sample_groups=dict(Car=20, Pedestrian=15, Cyclist=15))
+    sample_groups=dict(Car=20, Pedestrian=15, Cyclist=15),
+    points_loader=dict(
+        type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4))
 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),

configs/benchmark/hv_pointpillars_secfpn_3x8_100e_det3d_kitti-3d-car.py

@@ -83,7 +83,9 @@ db_sampler = dict(
     rate=1.0,
     prepare=dict(filter_by_difficulty=[-1], filter_by_min_points=dict(Car=5)),
     sample_groups=dict(Car=15),
-    classes=class_names)
+    classes=class_names,
+    points_loader=dict(
+        type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4))
 
 train_pipeline = [
     dict(type='LoadPointsFromFile', coord_type='LIDAR', load_dim=4, use_dim=4),