bugfixed: small bug in SPC sampling

pcdet/models/backbones_3d/pfe/voxel_set_abstraction.py

@@ -219,15 +219,15 @@ class VoxelSetAbstraction(nn.Module):
         Returns:
             sampled_points: (N_out, 3)
         """
+
         sampled_points, _ = sample_points_with_roi(
             rois=roi_boxes, points=points,
-            sample_radius_with_roi=self.model_cfg.SPC.SAMPLE_RADIUS_WITH_ROI,
-            num_max_points_of_part=self.model_cfg.SPC.get('NUM_POINTS_OF_EACH_SAMPLE_PART', 200000)
+            sample_radius_with_roi=self.model_cfg.SPC_SAMPLING.SAMPLE_RADIUS_WITH_ROI,
+            num_max_points_of_part=self.model_cfg.SPC_SAMPLING.get('NUM_POINTS_OF_EACH_SAMPLE_PART', 200000)
         )
-
         sampled_points = sector_fps(
             points=sampled_points, num_sampled_points=self.model_cfg.NUM_KEYPOINTS,
-            num_sectors=self.model_cfg.SPC.NUM_SECTORS
+            num_sectors=self.model_cfg.SPC_SAMPLING.NUM_SECTORS
         )
         return sampled_points
 
@@ -271,7 +271,7 @@ class VoxelSetAbstraction(nn.Module):
 
             elif self.model_cfg.SAMPLE_METHOD == 'SPC':
                 cur_keypoints = self.sectorized_proposal_centric_sampling(
-                    roi_boxes=batch_dict['rois'][bs_idx], points=sampled_points
+                    roi_boxes=batch_dict['rois'][bs_idx], points=sampled_points[0]
                 )
                 bs_idxs = cur_keypoints.new_ones(cur_keypoints.shape[0]) * bs_idx
                 keypoints = torch.cat((bs_idxs[:, None], cur_keypoints), dim=1)
@@ -290,7 +290,7 @@ class VoxelSetAbstraction(nn.Module):
     @staticmethod
     def aggregate_keypoint_features_from_one_source(
             batch_size, aggregate_func, xyz, xyz_features, xyz_bs_idxs, new_xyz, new_xyz_batch_cnt,
-            filter_neighbors_with_roi=False, radius_of_neighbor=None, num_max_points_of_part=None, rois=None
+            filter_neighbors_with_roi=False, radius_of_neighbor=None, num_max_points_of_part=200000, rois=None
     ):
         """
 
@@ -311,7 +311,7 @@ class VoxelSetAbstraction(nn.Module):
         """
         xyz_batch_cnt = xyz.new_zeros(batch_size).int()
         if filter_neighbors_with_roi:
-            point_features = torch.cat((xyz, xyz_features), dim=0) if xyz_features is not None else xyz
+            point_features = torch.cat((xyz, xyz_features), dim=-1) if xyz_features is not None else xyz
             point_features_list = []
             for bs_idx in range(batch_size):
                 bs_mask = (xyz_bs_idxs == bs_idx)
@@ -334,7 +334,7 @@ class VoxelSetAbstraction(nn.Module):
             xyz_batch_cnt=xyz_batch_cnt,
             new_xyz=new_xyz,
             new_xyz_batch_cnt=new_xyz_batch_cnt,
-            features=xyz_features,
+            features=xyz_features.contiguous(),
         )
         return pooled_features