You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello,
I want to test on a very small dataset.
I reduced my validation dataset size to only 3 point clouds. (The same issue is also present if dataset contains only 1 or 2 point clouds). The folder has the required structure - labels, images, calib, velodyne, velodyne reduced. create_data.py works fine. But test.py gives the following error. When I increase my data size, everything works fine again.
Traceback (most recent call last):
File "test.py", line 209, in
main()
File "test.py", line 204, in main
result = get_official_eval_result(gt_annos, outputs, current_classes=class_names)
File "/home/charvi/SA-SSD/mmdet/core/evaluation/kitti_eval.py", line 839, in get_official_eval_result
gt_annos, dt_annos, current_classes, min_overlaps, compute_aos, difficultys)
File "/home/charvi/SA-SSD/mmdet/core/evaluation/kitti_eval.py", line 708, in do_eval_v2
min_overlaps, compute_aos)
File "/home/charvi/SA-SSD/mmdet/core/evaluation/kitti_eval.py", line 586, in eval_class_v3
rets = calculate_iou_partly(dt_annos, gt_annos, metric, num_parts)
File "/home/charvi/SA-SSD/mmdet/core/evaluation/kitti_eval.py", line 376, in calculate_iou_partly
gt_boxes = np.concatenate([a["bbox"] for a in gt_annos_part], 0)
File "<array_function internals>", line 6, in concatenate
ValueError: need at least one array to concatenate
Any help is appreciated.
(In line 376 "gt_boxes = np.concatenate([a["bbox"] for a in gt_annos_part], 0)" gt_annos_part is coming an empty array)
Best regards,
Charvi
The text was updated successfully, but these errors were encountered:
CharviVitthal
changed the title
ValueError: need at least one array to concatenate
Unable to test on a very small dataset (ValueError: need at least one array to concatenate)
Oct 6, 2021
Hello,
I want to test on a very small dataset.
I reduced my validation dataset size to only 3 point clouds. (The same issue is also present if dataset contains only 1 or 2 point clouds). The folder has the required structure - labels, images, calib, velodyne, velodyne reduced. create_data.py works fine. But test.py gives the following error. When I increase my data size, everything works fine again.
Traceback (most recent call last):
File "test.py", line 209, in
main()
File "test.py", line 204, in main
result = get_official_eval_result(gt_annos, outputs, current_classes=class_names)
File "/home/charvi/SA-SSD/mmdet/core/evaluation/kitti_eval.py", line 839, in get_official_eval_result
gt_annos, dt_annos, current_classes, min_overlaps, compute_aos, difficultys)
File "/home/charvi/SA-SSD/mmdet/core/evaluation/kitti_eval.py", line 708, in do_eval_v2
min_overlaps, compute_aos)
File "/home/charvi/SA-SSD/mmdet/core/evaluation/kitti_eval.py", line 586, in eval_class_v3
rets = calculate_iou_partly(dt_annos, gt_annos, metric, num_parts)
File "/home/charvi/SA-SSD/mmdet/core/evaluation/kitti_eval.py", line 376, in calculate_iou_partly
gt_boxes = np.concatenate([a["bbox"] for a in gt_annos_part], 0)
File "<array_function internals>", line 6, in concatenate
ValueError: need at least one array to concatenate
Any help is appreciated.
(In line 376 "gt_boxes = np.concatenate([a["bbox"] for a in gt_annos_part], 0)" gt_annos_part is coming an empty array)
Best regards,
Charvi
The text was updated successfully, but these errors were encountered: