[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

tests/test_datasets/test_transforms/test_loading.py

@@ -80,10 +80,8 @@ class TestPointSegClassMapping(unittest.TestCase):
     def test_point_seg_class_mapping(self):
         results = dict()
         results['pts_semantic_mask'] = np.array([1, 2, 3, 4, 5])
-
-        point_seg_mapping_transform = PointSegClassMapping(
-            valid_cat_ids=[1, 2, 3],
-            max_cat_id=results['pts_semantic_mask'].max())
+        results['seg_label_mapping'] = np.array([3, 0, 1, 2, 3, 3])
+        point_seg_mapping_transform = PointSegClassMapping()
         results = point_seg_mapping_transform(results)
         assert_allclose(results['pts_semantic_mask'], np.array([0, 1, 2, 3,
                                                                 3]))

tests/test_evaluation/test_metrics/test_instance_seg_metric.py

@@ -67,14 +67,14 @@ class TestInstanceSegMetric(unittest.TestCase):
     def test_evaluate(self):
         data_batch = self._demo_mm_inputs()
         predictions = self._demo_mm_model_output()
-        valid_class_ids = (3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 16, 24, 28, 33,
-                           34, 36, 39)
+        seg_valid_class_ids = (3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 16, 24, 28,
+                               33, 34, 36, 39)
         class_labels = ('cabinet', 'bed', 'chair', 'sofa', 'table', 'door',
                         'window', 'bookshelf', 'picture', 'counter', 'desk',
                         'curtain', 'refrigerator', 'showercurtrain', 'toilet',
                         'sink', 'bathtub', 'garbagebin')
         dataset_meta = dict(
-            VALID_CLASS_IDS=valid_class_ids, CLASSES=class_labels)
+            seg_valid_class_ids=seg_valid_class_ids, CLASSES=class_labels)
         instance_seg_metric = InstanceSegMetric()
         instance_seg_metric.dataset_meta = dataset_meta
         instance_seg_metric.process(data_batch, predictions)

tests/test_evaluation/test_metrics/test_kitti_metric.py

@@ -20,6 +20,7 @@ def _init_evaluate_input():
     pred_instances_3d.labels_3d = torch.Tensor([0])
 
     predictions.pred_instances_3d = pred_instances_3d
+    predictions.pred_instances = InstanceData()
     predictions = predictions.to_dict()
     return data_batch, [predictions]
 

tests/test_evaluation/test_metrics/test_seg_eval.py

@@ -10,12 +10,12 @@ def test_indoor_eval():
     if not torch.cuda.is_available():
         pytest.skip()
     seg_preds = [
-        torch.Tensor([
+        np.array([
             0, 0, 1, 0, 0, 2, 1, 3, 1, 2, 1, 0, 2, 2, 2, 2, 1, 3, 0, 3, 3, 3, 3
         ])
     ]
     gt_labels = [
-        torch.Tensor([
+        np.array([
             0, 0, 0, 255, 0, 0, 1, 1, 1, 255, 1, 1, 2, 2, 2, 2, 2, 3, 3, 3, 3,
             3, 255
         ])

tests/test_evaluation/test_metrics/test_seg_metric.py

 # Copyright (c) OpenMMLab. All rights reserved.
 import unittest
 
+import numpy as np
 import torch
 from mmengine.data import BaseDataElement
 
@@ -14,7 +15,7 @@ class TestSegMetric(unittest.TestCase):
         """Create a superset of inputs needed to run test or train batches."""
         packed_inputs = []
         mm_inputs = dict()
-        pts_semantic_mask = torch.Tensor([
+        pts_semantic_mask = np.array([
             0, 0, 0, 255, 0, 0, 1, 1, 1, 255, 1, 1, 2, 2, 2, 2, 2, 3, 3, 3, 3,
             3, 255
         ])