[Refactor] move voxelization to data_preprocessor and fix new ut bugs (#1671)

* mv voxelization

* update

* update full

* fix configs

* improve docstring of data_preprocessor

* fix dynamic voxel config

* remove default voxel_type in config

* fix typos

* add docstring

* fix ut

* update

* fix docstring

configs/_base_/models/centerpoint_01voxel_second_secfpn_nus.py

 voxel_size = [0.1, 0.1, 0.2]
 model = dict(
     type='CenterPoint',
-    data_preprocessor=dict(type='Det3DDataPreprocessor'),
-    pts_voxel_layer=dict(
-        max_num_points=10, voxel_size=voxel_size, max_voxels=(90000, 120000)),
+    data_preprocessor=dict(
+        type='Det3DDataPreprocessor',
+        voxel=True,
+        voxel_layer=dict(
+            max_num_points=10,
+            voxel_size=voxel_size,
+            max_voxels=(90000, 120000))),
     pts_voxel_encoder=dict(type='HardSimpleVFE', num_features=5),
     pts_middle_encoder=dict(
         type='SparseEncoder',

configs/_base_/models/centerpoint_02pillar_second_secfpn_nus.py

 voxel_size = [0.2, 0.2, 8]
 model = dict(
     type='CenterPoint',
-    data_preprocessor=dict(type='Det3DDataPreprocessor'),
-    pts_voxel_layer=dict(
-        max_num_points=20, voxel_size=voxel_size, max_voxels=(30000, 40000)),
+    data_preprocessor=dict(
+        type='Det3DDataPreprocessor',
+        voxel=True,
+        voxel_layer=dict(
+            max_num_points=20,
+            voxel_size=voxel_size,
+            max_voxels=(30000, 40000))),
     pts_voxel_encoder=dict(
         type='PillarFeatureNet',
         in_channels=5,

configs/_base_/models/hv_pointpillars_fpn_lyft.py

@@ -6,10 +6,11 @@ _base_ = './hv_pointpillars_fpn_nus.py'
 # If point cloud range is modified, do remember to change all related
 # keys in the config.
 model = dict(
-    pts_voxel_layer=dict(
-        max_num_points=20,
-        point_cloud_range=[-80, -80, -5, 80, 80, 3],
-        max_voxels=(60000, 60000)),
+    data_preprocessor=dict(
+        voxel_layer=dict(
+            max_num_points=20,
+            point_cloud_range=[-80, -80, -5, 80, 80, 3],
+            max_voxels=(60000, 60000))),
     pts_voxel_encoder=dict(
         feat_channels=[64], point_cloud_range=[-80, -80, -5, 80, 80, 3]),
     pts_middle_encoder=dict(output_shape=[640, 640]),

configs/_base_/models/hv_pointpillars_fpn_nus.py

@@ -6,12 +6,14 @@
 voxel_size = [0.25, 0.25, 8]
 model = dict(
     type='MVXFasterRCNN',
-    data_preprocessor=dict(type='Det3DDataPreprocessor'),
-    pts_voxel_layer=dict(
-        max_num_points=64,
-        point_cloud_range=[-50, -50, -5, 50, 50, 3],
-        voxel_size=voxel_size,
-        max_voxels=(30000, 40000)),
+    data_preprocessor=dict(
+        type='Det3DDataPreprocessor',
+        voxel=True,
+        voxel_layer=dict(
+            max_num_points=64,
+            point_cloud_range=[-50, -50, -5, 50, 50, 3],
+            voxel_size=voxel_size,
+            max_voxels=(30000, 40000))),
     pts_voxel_encoder=dict(
         type='HardVFE',
         in_channels=4,

configs/_base_/models/hv_pointpillars_fpn_range100_lyft.py

@@ -6,10 +6,11 @@ _base_ = './hv_pointpillars_fpn_nus.py'
 # If point cloud range is modified, do remember to change all related
 # keys in the config.
 model = dict(
-    pts_voxel_layer=dict(
-        max_num_points=20,
-        point_cloud_range=[-100, -100, -5, 100, 100, 3],
-        max_voxels=(60000, 60000)),
+    data_preprocessor=dict(
+        voxel_layer=dict(
+            max_num_points=20,
+            point_cloud_range=[-100, -100, -5, 100, 100, 3],
+            max_voxels=(60000, 60000))),
     pts_voxel_encoder=dict(
         feat_channels=[64], point_cloud_range=[-100, -100, -5, 100, 100, 3]),
     pts_middle_encoder=dict(output_shape=[800, 800]),

configs/_base_/models/hv_pointpillars_secfpn_kitti.py

@@ -2,13 +2,14 @@ voxel_size = [0.16, 0.16, 4]
 
 model = dict(
     type='VoxelNet',
-    data_preprocessor=dict(type='Det3DDataPreprocessor'),
-    voxel_layer=dict(
-        max_num_points=32,  # max_points_per_voxel
-        point_cloud_range=[0, -39.68, -3, 69.12, 39.68, 1],
-        voxel_size=voxel_size,
-        max_voxels=(16000, 40000)  # (training, testing) max_voxels
-    ),
+    data_preprocessor=dict(
+        type='Det3DDataPreprocessor',
+        voxel=True,
+        voxel_layer=dict(
+            max_num_points=32,  # max_points_per_voxel
+            point_cloud_range=[0, -39.68, -3, 69.12, 39.68, 1],
+            voxel_size=voxel_size,
+            max_voxels=(16000, 40000))),
     voxel_encoder=dict(
         type='PillarFeatureNet',
         in_channels=4,

configs/_base_/models/hv_pointpillars_secfpn_waymo.py

@@ -6,12 +6,14 @@
 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],
-        voxel_size=voxel_size,
-        max_voxels=(32000, 32000)),
+    data_preprocessor=dict(
+        type='Det3DDataPreprocessor',
+        voxel=True,
+        voxel_layer=dict(
+            max_num_points=20,
+            point_cloud_range=[-74.88, -74.88, -2, 74.88, 74.88, 4],
+            voxel_size=voxel_size,
+            max_voxels=(32000, 32000))),
     pts_voxel_encoder=dict(
         type='HardVFE',
         in_channels=5,

configs/_base_/models/hv_second_secfpn_kitti.py

@@ -2,12 +2,14 @@ voxel_size = [0.05, 0.05, 0.1]
 
 model = dict(
     type='VoxelNet',
-    data_preprocessor=dict(type='Det3DDataPreprocessor'),
-    voxel_layer=dict(
-        max_num_points=5,
-        point_cloud_range=[0, -40, -3, 70.4, 40, 1],
-        voxel_size=voxel_size,
-        max_voxels=(16000, 40000)),
+    data_preprocessor=dict(
+        type='Det3DDataPreprocessor',
+        voxel=True,
+        voxel_layer=dict(
+            max_num_points=5,
+            point_cloud_range=[0, -40, -3, 70.4, 40, 1],
+            voxel_size=voxel_size,
+            max_voxels=(16000, 40000))),
     voxel_encoder=dict(type='HardSimpleVFE'),
     middle_encoder=dict(
         type='SparseEncoder',

configs/_base_/models/hv_second_secfpn_waymo.py

@@ -6,12 +6,14 @@
 voxel_size = [0.08, 0.08, 0.1]
 model = dict(
     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)),
+    data_preprocessor=dict(
+        type='Det3DDataPreprocessor',
+        voxel=True,
+        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))),
     pts_voxel_encoder=dict(type='HardSimpleVFE', num_features=5),
     pts_middle_encoder=dict(
         type='SparseEncoder',

configs/_base_/models/parta2.py

@@ -4,13 +4,14 @@ point_cloud_range = [0, -40, -3, 70.4, 40, 1]
 
 model = dict(
     type='PartA2',
-    data_preprocessor=dict(type='Det3DDataPreprocessor'),
-    voxel_layer=dict(
-        max_num_points=5,  # max_points_per_voxel
-        point_cloud_range=point_cloud_range,
-        voxel_size=voxel_size,
-        max_voxels=(16000, 40000)  # (training, testing) max_voxels
-    ),
+    data_preprocessor=dict(
+        type='Det3DDataPreprocessor',
+        voxel=True,
+        voxel_layer=dict(
+            max_num_points=5,  # max_points_per_voxel
+            point_cloud_range=point_cloud_range,
+            voxel_size=voxel_size,
+            max_voxels=(16000, 40000))),
     voxel_encoder=dict(type='HardSimpleVFE'),
     middle_encoder=dict(
         type='SparseUNet',

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

@@ -4,12 +4,14 @@ point_cloud_range = [0, -40, -3, 70.4, 40, 1]  # velodyne coordinates, x, y, z
 
 model = dict(
     type='PartA2',
-    voxel_layer=dict(
-        max_num_points=5,  # max_points_per_voxel
-        point_cloud_range=point_cloud_range,
-        voxel_size=voxel_size,
-        max_voxels=(16000, 40000)  # (training, testing) max_coxels
-    ),
+    data_preprocessor=dict(
+        type='Det3DDataPreprocessor',
+        voxel=True,
+        voxel_layer=dict(
+            max_num_points=5,  # max_points_per_voxel
+            point_cloud_range=point_cloud_range,
+            voxel_size=voxel_size,
+            max_voxels=(16000, 40000))),
     voxel_encoder=dict(type='HardSimpleVFE'),
     middle_encoder=dict(
         type='SparseUNet',

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

@@ -3,11 +3,14 @@ voxel_size = [0.16, 0.16, 4]
 point_cloud_range = [0, -39.68, -3, 69.12, 39.68, 1]
 model = dict(
     type='VoxelNet',
-    voxel_layer=dict(
-        max_num_points=64,
-        point_cloud_range=point_cloud_range,
-        voxel_size=voxel_size,
-        max_voxels=(12000, 20000)),
+    data_preprocessor=dict(
+        type='Det3DDataPreprocessor',
+        voxel=True,
+        voxel_layer=dict(
+            max_num_points=64,
+            point_cloud_range=point_cloud_range,
+            voxel_size=voxel_size,
+            max_voxels=(12000, 20000))),
     voxel_encoder=dict(
         type='PillarFeatureNet',
         in_channels=4,

configs/benchmark/hv_pointpillars_secfpn_4x8_80e_pcdet_kitti-3d-3class.py

@@ -3,12 +3,14 @@ point_cloud_range = [0, -39.68, -3, 69.12, 39.68, 1]
 voxel_size = [0.16, 0.16, 4]
 model = dict(
     type='VoxelNet',
-    voxel_layer=dict(
-        max_num_points=32,  # max_points_per_voxel
-        point_cloud_range=point_cloud_range,
-        voxel_size=voxel_size,
-        max_voxels=(16000, 40000)  # (training, testing) max_coxels
-    ),
+    data_preprocessor=dict(
+        type='Det3DDataPreprocessor',
+        voxel=True,
+        voxel_layer=dict(
+            max_num_points=32,  # max_points_per_voxel
+            point_cloud_range=point_cloud_range,
+            voxel_size=voxel_size,
+            max_voxels=(16000, 40000))),
     voxel_encoder=dict(
         type='PillarFeatureNet',
         in_channels=4,

configs/benchmark/hv_second_secfpn_4x8_80e_pcdet_kitti-3d-3class.py

@@ -4,11 +4,14 @@ point_cloud_range = [0, -40, -3, 70.4, 40, 1]
 
 model = dict(
     type='VoxelNet',
-    voxel_layer=dict(
-        max_num_points=5,
-        point_cloud_range=point_cloud_range,
-        voxel_size=voxel_size,
-        max_voxels=(16000, 40000)),
+    data_preprocessor=dict(
+        type='Det3DDataPreprocessor',
+        voxel=True,
+        voxel_layer=dict(
+            max_num_points=5,
+            point_cloud_range=point_cloud_range,
+            voxel_size=voxel_size,
+            max_voxels=(16000, 40000))),
     voxel_encoder=dict(type='HardSimpleVFE'),
     middle_encoder=dict(
         type='SparseEncoder',

configs/centerpoint/centerpoint_0075voxel_second_secfpn_4x8_cyclic_20e_nus.py

@@ -11,8 +11,9 @@ class_names = [
 ]
 data_prefix = dict(pts='samples/LIDAR_TOP', img='')
 model = dict(
-    pts_voxel_layer=dict(
-        voxel_size=voxel_size, point_cloud_range=point_cloud_range),
+    data_preprocessor=dict(
+        voxel_layer=dict(
+            voxel_size=voxel_size, point_cloud_range=point_cloud_range)),
     pts_middle_encoder=dict(sparse_shape=[41, 1440, 1440]),
     pts_bbox_head=dict(
         bbox_coder=dict(

configs/centerpoint/centerpoint_01voxel_second_secfpn_4x8_cyclic_20e_nus.py

@@ -14,7 +14,8 @@ class_names = [
 ]
 data_prefix = dict(pts='samples/LIDAR_TOP', img='')
 model = dict(
-    pts_voxel_layer=dict(point_cloud_range=point_cloud_range),
+    data_preprocessor=dict(
+        voxel_layer=dict(point_cloud_range=point_cloud_range)),
     pts_bbox_head=dict(bbox_coder=dict(pc_range=point_cloud_range[:2])),
     # model training and testing settings
     train_cfg=dict(pts=dict(point_cloud_range=point_cloud_range)),

configs/centerpoint/centerpoint_02pillar_second_secfpn_4x8_cyclic_20e_nus.py

@@ -14,7 +14,8 @@ class_names = [
 ]
 data_prefix = dict(pts='samples/LIDAR_TOP', img='')
 model = dict(
-    pts_voxel_layer=dict(point_cloud_range=point_cloud_range),
+    data_preprocessor=dict(
+        voxel_layer=dict(point_cloud_range=point_cloud_range)),
     pts_voxel_encoder=dict(point_cloud_range=point_cloud_range),
     pts_bbox_head=dict(bbox_coder=dict(pc_range=point_cloud_range[:2])),
     # model training and testing settings

configs/dynamic_voxelization/dv_pointpillars_secfpn_6x8_160e_kitti-3d-car.py

@@ -5,11 +5,13 @@ point_cloud_range = [0, -39.68, -3, 69.12, 39.68, 1]
 
 model = dict(
     type='DynamicVoxelNet',
-    voxel_layer=dict(
-        max_num_points=-1,
-        point_cloud_range=point_cloud_range,
-        voxel_size=voxel_size,
-        max_voxels=(-1, -1)),
+    data_preprocessor=dict(
+        voxel_type='dynamic',
+        voxel_layer=dict(
+            max_num_points=-1,
+            point_cloud_range=point_cloud_range,
+            voxel_size=voxel_size,
+            max_voxels=(-1, -1))),
     voxel_encoder=dict(
         type='DynamicPillarFeatureNet',
         in_channels=4,

configs/dynamic_voxelization/dv_second_secfpn_2x8_cosine_80e_kitti-3d-3class.py

@@ -9,12 +9,14 @@ voxel_size = [0.05, 0.05, 0.1]
 
 model = dict(
     type='DynamicVoxelNet',
-    voxel_layer=dict(
-        _delete_=True,
-        max_num_points=-1,
-        point_cloud_range=point_cloud_range,
-        voxel_size=voxel_size,
-        max_voxels=(-1, -1)),
+    data_preprocessor=dict(
+        voxel_type='dynamic',
+        voxel_layer=dict(
+            _delete_=True,
+            max_num_points=-1,
+            point_cloud_range=point_cloud_range,
+            voxel_size=voxel_size,
+            max_voxels=(-1, -1))),
     voxel_encoder=dict(
         _delete_=True,
         type='DynamicSimpleVFE',

configs/dynamic_voxelization/dv_second_secfpn_6x8_80e_kitti-3d-car.py

@@ -5,12 +5,14 @@ voxel_size = [0.05, 0.05, 0.1]
 
 model = dict(
     type='DynamicVoxelNet',
-    voxel_layer=dict(
-        _delete_=True,
-        max_num_points=-1,
-        point_cloud_range=point_cloud_range,
-        voxel_size=voxel_size,
-        max_voxels=(-1, -1)),
+    data_preprocessor=dict(
+        voxel_type='dynamic',
+        voxel_layer=dict(
+            _delete_=True,
+            max_num_points=-1,
+            point_cloud_range=point_cloud_range,
+            voxel_size=voxel_size,
+            max_voxels=(-1, -1))),
     voxel_encoder=dict(
         _delete_=True,
         type='DynamicSimpleVFE',