refactor directory

tests/test_data/test_datasets/test_scannet_dataset.py

@@ -5,8 +5,8 @@ import numpy as np
 import torch
 from mmengine.testing import assert_allclose
 
-from mmdet3d.core import DepthInstance3DBoxes
 from mmdet3d.datasets import ScanNetDataset, ScanNetSegDataset
+from mmdet3d.structures import DepthInstance3DBoxes
 from mmdet3d.utils import register_all_modules
 
 

tests/test_data/test_datasets/test_sunrgbd_dataset.py

@@ -5,8 +5,8 @@ import numpy as np
 import torch
 from mmengine.testing import assert_allclose
 
-from mmdet3d.core import DepthInstance3DBoxes
 from mmdet3d.datasets import SUNRGBDDataset
+from mmdet3d.structures import DepthInstance3DBoxes
 
 
 def _generate_scannet_dataset_config():

tests/test_data/test_transforms/test_augs.py

@@ -8,7 +8,7 @@ from mmengine.testing import assert_allclose
 from utils import create_data_info_after_loading
 
 from mmdet3d.datasets import GlobalAlignment, RandomFlip3D
-from mmdet3d.datasets.pipelines import GlobalRotScaleTrans
+from mmdet3d.datasets.transforms import GlobalRotScaleTrans
 
 
 class TestGlobalRotScaleTrans(unittest.TestCase):

tests/test_data/test_transforms/test_loading.py

@@ -6,10 +6,10 @@ import torch
 from mmengine.testing import assert_allclose
 from utils import create_dummy_data_info
 
-from mmdet3d.core import DepthPoints, LiDARPoints
-from mmdet3d.datasets.pipelines import PointSegClassMapping
-from mmdet3d.datasets.pipelines.loading import (LoadAnnotations3D,
+from mmdet3d.datasets.transforms import PointSegClassMapping
+from mmdet3d.datasets.transforms.loading import (LoadAnnotations3D,
                                                  LoadPointsFromFile)
+from mmdet3d.structures import DepthPoints, LiDARPoints
 
 
 class TestLoadPointsFromFile(unittest.TestCase):

tests/test_data/test_transforms/test_pack3d.py

@@ -5,8 +5,8 @@ import torch
 from mmengine.testing import assert_allclose
 from utils import create_data_info_after_loading
 
-from mmdet3d.core import LiDARInstance3DBoxes
-from mmdet3d.datasets.pipelines.formating import Pack3DDetInputs
+from mmdet3d.datasets.transforms.formating import Pack3DDetInputs
+from mmdet3d.structures import LiDARInstance3DBoxes
 
 
 class TestPack3DDetInputs(unittest.TestCase):

tests/test_data/test_transforms/utils.py

 # Copyright (c) OpenMMLab. All rights reserved.
 import numpy as np
 
-from mmdet3d.core import LiDARInstance3DBoxes
 # create a dummy `results` to test the pipeline
 from mmdet3d.datasets import LoadAnnotations3D, LoadPointsFromFile
-from mmdet3d.datasets.pipelines.loading import LoadImageFromFileMono3D
+from mmdet3d.datasets.transforms.loading import LoadImageFromFileMono3D
+from mmdet3d.structures import LiDARInstance3DBoxes
 
 
 def create_dummy_data_info(with_ann=True):

tests/test_metrics/test_indoor_eval.py

-# Copyright (c) OpenMMLab. All rights reserved.
-import numpy as np
-import pytest
-import torch
-
-from mmdet3d.core.evaluation.indoor_eval import average_precision, indoor_eval
-
-
-def test_indoor_eval():
-    if not torch.cuda.is_available():
-        pytest.skip()
-    from mmdet3d.core.bbox.structures import Box3DMode, DepthInstance3DBoxes
-    det_infos = [{
-        'labels_3d':
-        torch.tensor([0, 1, 2, 2, 0, 3, 1, 2, 3, 2]),
-        'boxes_3d':
-        DepthInstance3DBoxes(
-            torch.tensor([[
-                -2.4089e-03, -3.3174e+00, 4.9438e-01, 2.1668e+00, 2.8431e-01,
-                1.6506e+00, 0.0000e+00
-            ],
-                          [
-                              -3.4269e-01, -2.7565e+00, 2.8144e-02, 6.8554e-01,
-                              9.6854e-01, 6.1755e-01, 0.0000e+00
-                          ],
-                          [
-                              -3.8320e+00, -1.0646e+00, 1.7074e-01, 2.4981e-01,
-                              4.4708e-01, 6.2538e-01, 0.0000e+00
-                          ],
-                          [
-                              4.1073e-01, 3.3757e+00, 3.4311e-01, 8.0617e-01,
-                              2.8679e-01, 1.6060e+00, 0.0000e+00
-                          ],
-                          [
-                              6.1199e-01, -3.1041e+00, 4.1873e-01, 1.2310e+00,
-                              4.0162e-01, 1.7303e+00, 0.0000e+00
-                          ],
-                          [
-                              -5.9877e-01, -2.6011e+00, 1.1148e+00, 1.5704e-01,
-                              7.5957e-01, 9.6930e-01, 0.0000e+00
-                          ],
-                          [
-                              2.7462e-01, -3.0088e+00, 6.5231e-02, 8.1208e-01,
-                              4.1861e-01, 3.7339e-01, 0.0000e+00
-                          ],
-                          [
-                              -1.4704e+00, -2.0024e+00, 2.7479e-01, 1.7888e+00,
-                              1.0566e+00, 1.3704e+00, 0.0000e+00
-                          ],
-                          [
-                              8.2727e-02, -3.1160e+00, 2.5690e-01, 1.4054e+00,
-                              2.0772e-01, 9.6792e-01, 0.0000e+00
-                          ],
-                          [
-                              2.6896e+00, 1.9881e+00, 1.1566e+00, 9.9885e-02,
-                              3.5713e-01, 4.5638e-01, 0.0000e+00
-                          ]]),
-            origin=(0.5, 0.5, 0)),
-        'scores_3d':
-        torch.tensor([
-            1.7516e-05, 1.0167e-06, 8.4486e-07, 7.1048e-02, 6.4274e-05,
-            1.5003e-07, 5.8102e-06, 1.9399e-08, 5.3126e-07, 1.8630e-09
-        ])
-    }]
-
-    label2cat = {
-        0: 'cabinet',
-        1: 'bed',
-        2: 'chair',
-        3: 'sofa',
-    }
-    gt_annos = [{
-        'gt_num':
-        10,
-        'gt_boxes_upright_depth':
-        np.array([[
-            -2.4089e-03, -3.3174e+00, 4.9438e-01, 2.1668e+00, 2.8431e-01,
-            1.6506e+00, 0.0000e+00
-        ],
-                  [
-                      -3.4269e-01, -2.7565e+00, 2.8144e-02, 6.8554e-01,
-                      9.6854e-01, 6.1755e-01, 0.0000e+00
-                  ],
-                  [
-                      -3.8320e+00, -1.0646e+00, 1.7074e-01, 2.4981e-01,
-                      4.4708e-01, 6.2538e-01, 0.0000e+00
-                  ],
-                  [
-                      4.1073e-01, 3.3757e+00, 3.4311e-01, 8.0617e-01,
-                      2.8679e-01, 1.6060e+00, 0.0000e+00
-                  ],
-                  [
-                      6.1199e-01, -3.1041e+00, 4.1873e-01, 1.2310e+00,
-                      4.0162e-01, 1.7303e+00, 0.0000e+00
-                  ],
-                  [
-                      -5.9877e-01, -2.6011e+00, 1.1148e+00, 1.5704e-01,
-                      7.5957e-01, 9.6930e-01, 0.0000e+00
-                  ],
-                  [
-                      2.7462e-01, -3.0088e+00, 6.5231e-02, 8.1208e-01,
-                      4.1861e-01, 3.7339e-01, 0.0000e+00
-                  ],
-                  [
-                      -1.4704e+00, -2.0024e+00, 2.7479e-01, 1.7888e+00,
-                      1.0566e+00, 1.3704e+00, 0.0000e+00
-                  ],
-                  [
-                      8.2727e-02, -3.1160e+00, 2.5690e-01, 1.4054e+00,
-                      2.0772e-01, 9.6792e-01, 0.0000e+00
-                  ],
-                  [
-                      2.6896e+00, 1.9881e+00, 1.1566e+00, 9.9885e-02,
-                      3.5713e-01, 4.5638e-01, 0.0000e+00
-                  ]]),
-        'class':
-        np.array([0, 1, 2, 0, 0, 3, 1, 3, 3, 2])
-    }]
-
-    ret_value = indoor_eval(
-        gt_annos,
-        det_infos, [0.25, 0.5],
-        label2cat,
-        box_type_3d=DepthInstance3DBoxes,
-        box_mode_3d=Box3DMode.DEPTH)
-
-    assert np.isclose(ret_value['cabinet_AP_0.25'], 0.666667)
-    assert np.isclose(ret_value['bed_AP_0.25'], 1.0)
-    assert np.isclose(ret_value['chair_AP_0.25'], 0.5)
-    assert np.isclose(ret_value['mAP_0.25'], 0.708333)
-    assert np.isclose(ret_value['mAR_0.25'], 0.833333)
-
-
-def test_indoor_eval_less_classes():
-    if not torch.cuda.is_available():
-        pytest.skip()
-    from mmdet3d.core.bbox.structures import Box3DMode, DepthInstance3DBoxes
-    det_infos = [{
-        'labels_3d':
-        torch.tensor([0]),
-        'boxes_3d':
-        DepthInstance3DBoxes(torch.tensor([[1., 1., 1., 1., 1., 1., 1.]])),
-        'scores_3d':
-        torch.tensor([.5])
-    }, {
-        'labels_3d':
-        torch.tensor([1]),
-        'boxes_3d':
-        DepthInstance3DBoxes(torch.tensor([[1., 1., 1., 1., 1., 1., 1.]])),
-        'scores_3d':
-        torch.tensor([.5])
-    }]
-
-    label2cat = {0: 'cabinet', 1: 'bed', 2: 'chair'}
-    gt_annos = [{
-        'gt_num':
-        2,
-        'gt_boxes_upright_depth':
-        np.array([[0., 0., 0., 1., 1., 1., 1.], [1., 1., 1., 1., 1., 1., 1.]]),
-        'class':
-        np.array([2, 0])
-    }, {
-        'gt_num':
-        1,
-        'gt_boxes_upright_depth':
-        np.array([
-            [1., 1., 1., 1., 1., 1., 1.],
-        ]),
-        'class':
-        np.array([1])
-    }]
-
-    ret_value = indoor_eval(
-        gt_annos,
-        det_infos, [0.25, 0.5],
-        label2cat,
-        box_type_3d=DepthInstance3DBoxes,
-        box_mode_3d=Box3DMode.DEPTH)
-
-    assert np.isclose(ret_value['mAP_0.25'], 0.666667)
-    assert np.isclose(ret_value['mAR_0.25'], 0.666667)
-
-
-def test_average_precision():
-    ap = average_precision(
-        np.array([[0.25, 0.5, 0.75], [0.25, 0.5, 0.75]]),
-        np.array([[1., 1., 1.], [1., 1., 1.]]), '11points')
-    assert abs(ap[0] - 0.06611571) < 0.001

tests/test_metrics/test_indoor_metric.py

@@ -5,8 +5,8 @@ from unittest.mock import patch
 import numpy as np
 import torch
 
-from mmdet3d.core import DepthInstance3DBoxes
-from mmdet3d.metrics import IndoorMetric
+from mmdet3d.evaluation.metrics import IndoorMetric
+from mmdet3d.structures import DepthInstance3DBoxes
 
 
 class TestIndoorMetric(unittest.TestCase):

tests/test_metrics/test_instance_seg_eval.py

@@ -2,7 +2,7 @@
 import numpy as np
 import torch
 
-from mmdet3d.core import instance_seg_eval
+from mmdet3d.evaluation import instance_seg_eval
 
 
 def test_instance_seg_eval():

tests/test_metrics/test_instance_seg_metric.py

@@ -5,8 +5,8 @@ import numpy as np
 import torch
 from mmengine.data import BaseDataElement
 
-from mmdet3d.core.data_structures import Det3DDataSample, PointData
-from mmdet3d.metrics import InstanceSegMetric
+from mmdet3d.evaluation.metrics import InstanceSegMetric
+from mmdet3d.structures import Det3DDataSample, PointData
 
 
 class TestInstanceSegMetric(unittest.TestCase):

tests/test_metrics/test_kitti_eval.py

@@ -3,8 +3,7 @@ import numpy as np
 import pytest
 import torch
 
-from mmdet3d.core.evaluation.kitti_utils.eval import (do_eval, eval_class,
-                                                      kitti_eval)
+from mmdet3d.evaluation import do_eval, eval_class, kitti_eval
 
 
 def test_do_eval():

tests/test_metrics/test_kitti_metric.py

@@ -3,9 +3,8 @@ import pytest
 import torch
 from mmengine.data import InstanceData
 
-from mmdet3d.core import Det3DDataSample
-from mmdet3d.core.bbox import LiDARInstance3DBoxes
-from mmdet3d.metrics import KittiMetric
+from mmdet3d.evaluation.metrics import KittiMetric
+from mmdet3d.structures import Det3DDataSample, LiDARInstance3DBoxes
 
 data_root = 'tests/data/kitti'
 

tests/test_metrics/test_seg_eval.py

@@ -3,7 +3,7 @@ import numpy as np
 import pytest
 import torch
 
-from mmdet3d.core.evaluation.seg_eval import seg_eval
+from mmdet3d.evaluation.functional.seg_eval import seg_eval
 
 
 def test_indoor_eval():

tests/test_metrics/test_seg_metric.py

@@ -4,8 +4,8 @@ import unittest
 import torch
 from mmengine.data import BaseDataElement
 
-from mmdet3d.core.data_structures import Det3DDataSample, PointData
-from mmdet3d.metrics import SegMetric
+from mmdet3d.evaluation.metrics import SegMetric
+from mmdet3d.structures import Det3DDataSample, PointData
 
 
 class TestSegMetric(unittest.TestCase):

tests/test_models/test_common_modules/test_dgcnn_modules.py

@@ -7,7 +7,7 @@ import torch
 def test_dgcnn_gf_module():
     if not torch.cuda.is_available():
         pytest.skip()
-    from mmdet3d.ops import DGCNNGFModule
+    from mmdet3d.models.layers import DGCNNGFModule
 
     self = DGCNNGFModule(
         mlp_channels=[18, 64, 64],
@@ -60,7 +60,7 @@ def test_dgcnn_gf_module():
 def test_dgcnn_fa_module():
     if not torch.cuda.is_available():
         pytest.skip()
-    from mmdet3d.ops import DGCNNFAModule
+    from mmdet3d.models.layers import DGCNNFAModule
 
     self = DGCNNFAModule(mlp_channels=[24, 16]).cuda()
     assert self.mlps.layer0.conv.in_channels == 24
@@ -75,7 +75,7 @@ def test_dgcnn_fa_module():
 def test_dgcnn_fp_module():
     if not torch.cuda.is_available():
         pytest.skip()
-    from mmdet3d.ops import DGCNNFPModule
+    from mmdet3d.models.layers import DGCNNFPModule
 
     self = DGCNNFPModule(mlp_channels=[24, 16]).cuda()
     assert self.mlps.layer0.conv.in_channels == 24

tests/test_models/test_common_modules/test_paconv_modules.py

@@ -7,7 +7,7 @@ import torch
 def test_paconv_sa_module_msg():
     if not torch.cuda.is_available():
         pytest.skip()
-    from mmdet3d.ops import PAConvSAModuleMSG
+    from mmdet3d.models.layers import PAConvSAModuleMSG
 
     # paconv_num_kernels should have same length as mlp_channels
     with pytest.raises(AssertionError):
@@ -105,7 +105,7 @@ def test_paconv_sa_module_msg():
 def test_paconv_sa_module():
     if not torch.cuda.is_available():
         pytest.skip()
-    from mmdet3d.ops import build_sa_module
+    from mmdet3d.models.layers import build_sa_module
     sa_cfg = dict(
         type='PAConvSAModule',
         num_point=16,
@@ -164,7 +164,7 @@ def test_paconv_sa_module():
 def test_paconv_cuda_sa_module_msg():
     if not torch.cuda.is_available():
         pytest.skip()
-    from mmdet3d.ops import PAConvCUDASAModuleMSG
+    from mmdet3d.models.layers import PAConvCUDASAModuleMSG
 
     # paconv_num_kernels should have same length as mlp_channels
     with pytest.raises(AssertionError):
@@ -245,7 +245,7 @@ def test_paconv_cuda_sa_module_msg():
 def test_paconv_cuda_sa_module():
     if not torch.cuda.is_available():
         pytest.skip()
-    from mmdet3d.ops import build_sa_module
+    from mmdet3d.models.layers import build_sa_module
     sa_cfg = dict(
         type='PAConvCUDASAModule',
         num_point=16,

tests/test_models/test_common_modules/test_paconv_ops.py

@@ -2,7 +2,7 @@
 import pytest
 import torch
 
-from mmdet3d.ops import PAConv, PAConvCUDA
+from mmdet3d.models.layers import PAConv, PAConvCUDA
 
 
 def test_paconv():

tests/test_models/test_common_modules/test_pointnet_modules.py

@@ -7,7 +7,7 @@ import torch
 def test_pointnet_sa_module_msg():
     if not torch.cuda.is_available():
         pytest.skip()
-    from mmdet3d.ops import PointSAModuleMSG
+    from mmdet3d.models.layers import PointSAModuleMSG
 
     self = PointSAModuleMSG(
         num_point=16,
@@ -152,7 +152,7 @@ def test_pointnet_sa_module_msg():
 def test_pointnet_sa_module():
     if not torch.cuda.is_available():
         pytest.skip()
-    from mmdet3d.ops import build_sa_module
+    from mmdet3d.models.layers import build_sa_module
     sa_cfg = dict(
         type='PointSAModule',
         num_point=16,
@@ -211,7 +211,7 @@ def test_pointnet_sa_module():
 def test_pointnet_fp_module():
     if not torch.cuda.is_available():
         pytest.skip()
-    from mmdet3d.ops import PointFPModule
+    from mmdet3d.models.layers import PointFPModule
 
     self = PointFPModule(mlp_channels=[24, 16]).cuda()
     assert self.mlps.layer0.conv.in_channels == 24

tests/test_models/test_common_modules/test_sparse_unet.py

@@ -2,8 +2,8 @@
 import pytest
 import torch
 
-from mmdet3d.ops import SparseBasicBlock
-from mmdet3d.ops.spconv import IS_SPCONV2_AVAILABLE
+from mmdet3d.models.layers import SparseBasicBlock
+from mmdet3d.models.layers.spconv import IS_SPCONV2_AVAILABLE
 
 if IS_SPCONV2_AVAILABLE:
     from spconv.pytorch import (SparseConv3d, SparseConvTensor,
@@ -97,7 +97,7 @@ def test_SparseBasicBlock():
 def test_make_sparse_convmodule():
     if not torch.cuda.is_available():
         pytest.skip('test requires GPU and torch+cuda')
-    from mmdet3d.ops import make_sparse_convmodule
+    from mmdet3d.models.layers import make_sparse_convmodule
 
     voxel_features = torch.tensor(
         [[6.56126, 0.9648336, -1.7339306, 0.315],

tests/test_models/test_common_modules/test_vote_module.py

@@ -3,7 +3,7 @@ import torch
 
 
 def test_vote_module():
-    from mmdet3d.models.model_utils import VoteModule
+    from mmdet3d.models.layers import VoteModule
 
     vote_loss = dict(
         type='ChamferDistance',