[Fix]: fix multi-batch test problem (#95)

* update bbox_results to support multi-batches test and fix ci problem in test_part_aggregation_ROI_head

* compress test samples in test_parta2_roi_head

* compress test samples

* fix multi-batch bug

* fix multi-batch bug

* fix multi-batch bug

* fix mvxnet test

* fix mvxnet test

* multi batch test

* multi batch test

* multi batch test

mmdet3d/apis/test.py

@@ -31,7 +31,7 @@ def single_gpu_test(model, data_loader, show=False, out_dir=None):
         if show:
             model.module.show_results(data, result, out_dir)
 
-        results.append(result)
+        results.extend(result)
 
         batch_size = len(data['img_metas'][0].data)
         for _ in range(batch_size):

mmdet3d/models/detectors/base.py

@@ -35,9 +35,6 @@ class Base3DDetector(BaseDetector):
             raise ValueError(
                 'num of augmentations ({}) != num of image meta ({})'.format(
                     len(points), len(img_metas)))
-        # TODO: remove the restriction of imgs_per_gpu == 1 when prepared
-        samples_per_gpu = len(points[0])
-        assert samples_per_gpu == 1
 
         if num_augs == 1:
             img = [img] if img is None else img

mmdet3d/models/detectors/h3dnet.py

@@ -163,7 +163,7 @@ class H3DNet(TwoStage3DDetector):
         merged_bboxes = merge_aug_bboxes_3d(aug_bboxes, img_metas,
                                             self.bbox_head.test_cfg)
 
-        return merged_bboxes
+        return [merged_bboxes]
 
     def extract_feats(self, points, img_metas):
         """Extract features of multiple samples."""

mmdet3d/models/detectors/mvx_two_stage.py

@@ -389,22 +389,24 @@ class MVXTwoStageDetector(Base3DDetector):
             bbox3d2result(bboxes, scores, labels)
             for bboxes, scores, labels in bbox_list
         ]
-        return bbox_results[0]
+        return bbox_results
 
     def simple_test(self, points, img_metas, img=None, rescale=False):
         """Test function without augmentaiton."""
         img_feats, pts_feats = self.extract_feat(
             points, img=img, img_metas=img_metas)
 
-        bbox_list = dict()
+        bbox_list = [dict() for i in range(len(img_metas))]
         if pts_feats and self.with_pts_bbox:
             bbox_pts = self.simple_test_pts(
                 pts_feats, img_metas, rescale=rescale)
-            bbox_list.update(pts_bbox=bbox_pts)
+            for result_dict, pts_bbox in zip(bbox_list, bbox_pts):
+                result_dict['pts_bbox'] = pts_bbox
         if img_feats and self.with_img_bbox:
             bbox_img = self.simple_test_img(
                 img_feats, img_metas, rescale=rescale)
-            bbox_list.update(img_bbox=bbox_img)
+            for result_dict, img_bbox in zip(bbox_list, bbox_img):
+                result_dict['img_bbox'] = img_bbox
         return bbox_list
 
     def aug_test(self, points, img_metas, imgs=None, rescale=False):
@@ -415,7 +417,7 @@ class MVXTwoStageDetector(Base3DDetector):
         if pts_feats and self.with_pts_bbox:
             bbox_pts = self.aug_test_pts(pts_feats, img_metas, rescale)
             bbox_list.update(pts_bbox=bbox_pts)
-        return bbox_list
+        return [bbox_list]
 
     def extract_feats(self, points, img_metas, imgs=None):
         """Extract point and image features of multiple samples."""

mmdet3d/models/detectors/votenet.py

@@ -81,7 +81,7 @@ class VoteNet(SingleStage3DDetector):
             bbox3d2result(bboxes, scores, labels)
             for bboxes, scores, labels in bbox_list
         ]
-        return bbox_results[0]
+        return bbox_results
 
     def aug_test(self, points, img_metas, imgs=None, rescale=False):
         """Test with augmentation."""
@@ -104,4 +104,4 @@ class VoteNet(SingleStage3DDetector):
         merged_bboxes = merge_aug_bboxes_3d(aug_bboxes, img_metas,
                                             self.bbox_head.test_cfg)
 
-        return merged_bboxes
+        return [merged_bboxes]

mmdet3d/models/detectors/voxelnet.py

@@ -101,7 +101,7 @@ class VoxelNet(SingleStage3DDetector):
             bbox3d2result(bboxes, scores, labels)
             for bboxes, scores, labels in bbox_list
         ]
-        return bbox_results[0]
+        return bbox_results
 
     def aug_test(self, points, img_metas, imgs=None, rescale=False):
         """Test function with augmentaiton."""
@@ -123,4 +123,4 @@ class VoxelNet(SingleStage3DDetector):
         merged_bboxes = merge_aug_bboxes_3d(aug_bboxes, img_metas,
                                             self.bbox_head.test_cfg)
 
-        return merged_bboxes
+        return [merged_bboxes]

mmdet3d/models/roi_heads/h3d_roi_head.py

@@ -155,4 +155,4 @@ class H3DRoIHead(Base3DRoIHead):
             bbox3d2result(bboxes, scores, labels)
             for bboxes, scores, labels in bbox_list
         ]
-        return bbox_results[0]
+        return bbox_results

mmdet3d/models/roi_heads/part_aggregation_roi_head.py

@@ -159,7 +159,7 @@ class PartAggregationROIHead(Base3DRoIHead):
             bbox3d2result(bboxes, scores, labels)
             for bboxes, scores, labels in bbox_list
         ]
-        return bbox_results[0]
+        return bbox_results
 
     def _bbox_forward_train(self, seg_feats, part_feats, voxels_dict,
                             sampling_results):

tests/test_apis.py

@@ -38,9 +38,9 @@ def test_inference_detector():
                    '6x8_160e_kitti-3d-3class.py'
     detector = init_detector(detector_cfg, device='cpu')
     results = inference_detector(detector, pcd)
-    bboxes_3d = results[0]['boxes_3d']
-    scores_3d = results[0]['scores_3d']
-    labels_3d = results[0]['labels_3d']
+    bboxes_3d = results[0][0]['boxes_3d']
+    scores_3d = results[0][0]['scores_3d']
+    labels_3d = results[0][0]['labels_3d']
     assert bboxes_3d.tensor.shape[0] >= 0
     assert bboxes_3d.tensor.shape[1] == 7
     assert scores_3d.shape[0] >= 0

tests/test_detectors.py

@@ -113,9 +113,9 @@ def test_voxel_net():
 
     # test simple_test
     results = self.simple_test(points, img_metas)
-    boxes_3d = results['boxes_3d']
-    scores_3d = results['scores_3d']
-    labels_3d = results['labels_3d']
+    boxes_3d = results[0]['boxes_3d']
+    scores_3d = results[0]['scores_3d']
+    labels_3d = results[0]['labels_3d']
     assert boxes_3d.tensor.shape == (50, 7)
     assert scores_3d.shape == torch.Size([50])
     assert labels_3d.shape == torch.Size([50])
@@ -155,9 +155,9 @@ def test_vote_net():
 
     # test simple_test
     results = self.simple_test(points, img_metas)
-    boxes_3d = results['boxes_3d']
-    scores_3d = results['scores_3d']
-    labels_3d = results['labels_3d']
+    boxes_3d = results[0]['boxes_3d']
+    scores_3d = results[0]['scores_3d']
+    labels_3d = results[0]['labels_3d']
     assert boxes_3d.tensor.shape[0] >= 0
     assert boxes_3d.tensor.shape[1] == 7
     assert scores_3d.shape[0] >= 0
@@ -171,8 +171,8 @@ def test_parta2():
     parta2 = _get_detector_cfg(
         'parta2/hv_PartA2_secfpn_2x8_cyclic_80e_kitti-3d-3class.py')
     self = build_detector(parta2).cuda()
-    points_0 = torch.rand([2000, 4], device='cuda')
-    points_1 = torch.rand([2000, 4], device='cuda')
+    points_0 = torch.rand([1000, 4], device='cuda')
+    points_1 = torch.rand([1000, 4], device='cuda')
     points = [points_0, points_1]
     img_meta_0 = dict(box_type_3d=LiDARInstance3DBoxes)
     img_meta_1 = dict(box_type_3d=LiDARInstance3DBoxes)
@@ -197,9 +197,9 @@ def test_parta2():
 
     # test_simple_test
     results = self.simple_test(points, img_metas)
-    boxes_3d = results['boxes_3d']
-    scores_3d = results['scores_3d']
-    labels_3d = results['labels_3d']
+    boxes_3d = results[0]['boxes_3d']
+    scores_3d = results[0]['scores_3d']
+    labels_3d = results[0]['labels_3d']
     assert boxes_3d.tensor.shape[0] >= 0
     assert boxes_3d.tensor.shape[1] == 7
     assert scores_3d.shape[0] >= 0

tests/test_heads.py

@@ -372,22 +372,20 @@ def test_part_aggregation_ROI_head():
     if not torch.cuda.is_available():
         pytest.skip('test requires GPU and torch+cuda')
 
-    _setup_seed(0)
     roi_head_cfg = _get_roi_head_cfg(
         'parta2/hv_PartA2_secfpn_2x8_cyclic_80e_kitti-3d-3class.py')
     self = build_head(roi_head_cfg).cuda()
-    spatial_features = torch.rand([1, 256, 200, 176], device='cuda')
-    seg_features = torch.rand([32000, 16], device='cuda')
-    neck_features = [torch.rand([1, 512, 200, 176], device='cuda')]
-    feats_dict = dict(
-        spatial_features=spatial_features,
-        seg_features=seg_features,
-        neck_features=neck_features)
-
-    voxels = torch.rand([32000, 5, 4], device='cuda')
-    num_points = torch.ones([32000], device='cuda')
-    coors = torch.zeros([32000, 4], device='cuda')
-    voxel_centers = torch.zeros([32000, 3], device='cuda')
+
+    features = np.load('./tests/test_samples/parta2_roihead_inputs.npz')
+    seg_features = torch.tensor(
+        features['seg_features'], dtype=torch.float32, device='cuda')
+    feats_dict = dict(seg_features=seg_features)
+
+    voxels = torch.tensor(
+        features['voxels'], dtype=torch.float32, device='cuda')
+    num_points = torch.ones([500], device='cuda')
+    coors = torch.zeros([500, 4], device='cuda')
+    voxel_centers = torch.zeros([500, 3], device='cuda')
     box_type_3d = LiDARInstance3DBoxes
     img_metas = [dict(box_type_3d=box_type_3d)]
     voxels_dict = dict(
@@ -396,10 +394,27 @@ def test_part_aggregation_ROI_head():
         coors=coors,
         voxel_centers=voxel_centers)
 
-    pred_bboxes = LiDARInstance3DBoxes(torch.rand([5, 7], device='cuda'))
-    pred_scores = torch.rand([5], device='cuda')
-    pred_labels = torch.randint(0, 3, [5], device='cuda')
-    pred_clses = torch.rand([5, 3], device='cuda')
+    pred_bboxes = LiDARInstance3DBoxes(
+        torch.tensor(
+            [[0.3990, 0.5167, 0.0249, 0.9401, 0.9459, 0.7967, 0.4150],
+             [0.8203, 0.2290, 0.9096, 0.1183, 0.0752, 0.4092, 0.9601],
+             [0.2093, 0.1940, 0.8909, 0.4387, 0.3570, 0.5454, 0.8299],
+             [0.2099, 0.7684, 0.4290, 0.2117, 0.6606, 0.1654, 0.4250],
+             [0.9927, 0.6964, 0.2472, 0.7028, 0.7494, 0.9303, 0.0494]],
+            dtype=torch.float32,
+            device='cuda'))
+    pred_scores = torch.tensor([0.9722, 0.7910, 0.4690, 0.3300, 0.3345],
+                               dtype=torch.float32,
+                               device='cuda')
+    pred_labels = torch.tensor([0, 1, 0, 2, 1],
+                               dtype=torch.int64,
+                               device='cuda')
+    pred_clses = torch.tensor(
+        [[0.7874, 0.1344, 0.2190], [0.8193, 0.6969, 0.7304],
+         [0.2328, 0.9028, 0.3900], [0.6177, 0.5012, 0.2330],
+         [0.8985, 0.4894, 0.7152]],
+        dtype=torch.float32,
+        device='cuda')
     proposal = dict(
         boxes_3d=pred_bboxes,
         scores_3d=pred_scores,
@@ -419,12 +434,12 @@ def test_part_aggregation_ROI_head():
 
     bbox_results = self.simple_test(feats_dict, voxels_dict, img_metas,
                                     proposal_list)
-    boxes_3d = bbox_results['boxes_3d']
-    scores_3d = bbox_results['scores_3d']
-    labels_3d = bbox_results['labels_3d']
-    assert boxes_3d.tensor.shape == (6, 7)
-    assert scores_3d.shape == (6, )
-    assert labels_3d.shape == (6, )
+    boxes_3d = bbox_results[0]['boxes_3d']
+    scores_3d = bbox_results[0]['scores_3d']
+    labels_3d = bbox_results[0]['labels_3d']
+    assert boxes_3d.tensor.shape == (12, 7)
+    assert scores_3d.shape == (12, )
+    assert labels_3d.shape == (12, )
 
 
 def test_free_anchor_3D_head():

tools/test.py

@@ -101,11 +101,11 @@ def main():
         set_random_seed(args.seed, deterministic=args.deterministic)
 
     # build the dataloader
-    # TODO: support multiple images per gpu (only minor changes are needed)
+    samples_per_gpu = cfg.data.test.pop('samples_per_gpu', 1)
     dataset = build_dataset(cfg.data.test)
     data_loader = build_dataloader(
         dataset,
-        samples_per_gpu=1,
+        samples_per_gpu=samples_per_gpu,
         workers_per_gpu=cfg.data.workers_per_gpu,
         dist=distributed,
         shuffle=False)